The problem with ECM isn't just that it's too good; it's that it is the defining component of one of this game's four pillars, notably information warfare, and that as a gameplay mechanic, it provides very little depth the way it is currently structured. It's an on/off cloaking switch, and the only way it interacts with every other component of electronic warfare in this game is being something to turn off. How do ECM and TAG interact? TAG cancels ECM. Same with NARC, BAP, and PPCs. This limits the possible depth of electronic warfare in this game to, "Am I cloaked to radar?" This is a relatively fulfilling state of affairs, as we'd been hoping for extra sensors, the ability to designate targets for teammates, multi-targeting functions, and more. Instead we have small bubbles of radar cloaking, and nothing else.
This also ties into a bigger, underling problem- specifically that without ewar capabilities, lights are going to remain perpetually hampered. There is a dilemma- should lights be equal to heavies in combat capability? If no, then there's very little reason to take a light in the game as it exists now. If yes, then why do lights cost so much more, and how should they be made equal to mechs with much more armor and weaponry? Either answer is problematic. The solution is to give lights and mediums a way to be useful without making them direct monsters in combat. If information warfare is fixed, role warfare will follow.
The first thing that needs to be done to address these issues is to make ECM actually function like a countermeasure, which is what it is supposed to be. Right now, ECM serves primarily an offensive function- it hides targets on radar, allowing them to fire without being detected in time for return fire, and allows lances to push in partial cover (i.e. treelines) without being detected. BAP functions as a defensive piece of equipment, and is mostly used to counter ECM when in range, and increase the distance at which enemy pushes can be spotted. This is backwards- BAP should provide a bonus that is countered by ECM, rather than the other way around!
As such, BAP is the first thing on the table for an overhaul; it needs to be something worth taking on its own merits for people who are scouting, rather than a tool for Warhawks and Stalkers to get slightly faster lock times. A good place to start is to look to the lore- BAP should allow limited targeting through terrain and obstacles. Give BAP a 250m radius of detection through obstacles and let the user target mechs within that radius without line of sight- but let the targets show up as unknown hostiles. This makes BAP ideal for people who are keen to scout enemy positions, but discourages taking it on every single mech, as it provides no information on what kinds of enemies you will encounter, what components are damaged, etc. If you think this sounds like a powerful ability, you're right- but it's not much more powerful than the already-existing seismic sensor, which does the same thing but for no tonnage cost or crit slots, unlike BAP. The one advantage that BAP has is allowing the mech scouting to share this intel with the team.
BAP should also allow for the targeting of multiple mechs at the same time; three sounds reasonable, but the number is definitely something up for debate, and is subject to the skill overhaul that this post covers later. It would function much the same way as the through-object targeting, where all targets beyond the first have their location marked for the spotter's allies, but no mech info is shared, and all non-primary targets show up as unknown, and LRMs targeting such non-primary targets would take an accuracy/tracking penalty. Secondary, tertiary, etc targets would be automatically added every time the spotter targets a new mech, meaning that the r key targets mechs as normal, and every time the target is switched the old primary target becomes the secondary, etc. All other normal targeting rules apply, i.e. when line of sight is lost, targets become unavailable after a short delay. Allowing mechs to target multiple hostiles is critical for teamwork in a game with limited communication functions like MWO; it lets a single spotter provide information on the movement a large group of enemies without having to be on TeamSpeak. This also adds incentives for enemies to run ECM and stay with an ewar mech.
Similarly, TAG and NARC need tweaking- they should both be pieces of equipment that provide bonuses to information gathering, rather than yet more ways to counter ECM. Both are relatively useful in their current state, and should continue to provide missile tracking strength buffs. But they also deserve to have a non-missile use, so that scout/ewar mechs equipped with them do not have dead tonnage should their teammates not have LRMs equipped. As such, make NARCed mechs instantly broadcast their health data, and TAGged mechs have their data gathered 50% faster. Depending on balance needs, TAG could be made invisible in normal vision modes, and glow brightly in thermal, as it is supposed to be an infrared beam.
Because of how powerful this makes BAP, TAG, and NARC, now we actually have somethings that need a countermeasures- this is where ECM steps in. ECM now shields units from being detected by BAP's non-LOS targeting, and nullifies the missile tracking strength buff given by TAG and NARC. It continues to provide decreased visibility to normal sensors as it does now- but units shielded by ECM should be visible on unaugmented sensors at 5-600m, which provides teams being pushed by ECM lances enough time to engage with missiles for a limited window before the enemy closes distance, yet still provides a distinct advantage to a defending team being bombarded by LRMs. Furthermore, the ECM-LRM interaction needs to change- ECM being a hard counter to LRMs by making locks impossible or making locks so difficult to obtain it might as well be impossible makes for unfun gameplay for LRM users, and means that LRMs need to be balanced such that they are very strong against unshielded teams because of the risk when engaging against an ECM-shielded team. Therefore, remove the LRM-lock penalty provided by ECM, and instead give LRMs a tracking strength penalty when targeting ECM-covered mechs.
The command console is finally also not useless, but it doesn't give nearly as much benefit as the clan targeting computers, for significantly more weight. I propose to give it utility by restricting airstrikes and artillery to mechs with command consoles; additional functionality could include calling down airstrikes from the battlegrid on a command console mech, and allowing a mech that both has lance command and a command console to designate targets with a star on their lance's IFF to coordinate fire.
There are still a few pieces of equiment left to cover, but I'll get to those. For the most part, that covers the basics of ewar, and if nothing but the above were to be implemented, it'd be a good start. Things really get interesting when you consider the ramifications for role warfare that having a viable scouting role would give- meaning, an overhaul to the pilot tree would do a world of good to compliment the electronics overhaul.
To kick things off, the pilot tree needs to be, well, a tree. What we have now is linear- it offers no choice, every skill is always eventually unlocked, there's no differentiation. Instead, a tree-structure should have, well, branches, different paths that diverge and let the pilot specialize in specific aspects of combat on a per-chassis basis, rather than every mech getting all the same skills.
The three groups I propose are Recon, Movement and Combat- it will be possible to go into more than one tree, but only one can attain master rank. Why these three? Because combined, they cover almost every possible archetype of mech that exists in MWO, and in a game with as rich of customization as MWO, flexibility is important.
Take the light class, for example. A Locust, with few weapons but excellent speed, would benefit immensely from having movement and recon abilities. A Raven might want offense for its large laser combined with recon. A Jenner might go for movement and combat. And these skill archetypes go all the way up the tree- a Victor would probably take movemement and combat buffs, while an D-DC would go for ewar and firepower boosts.
An overhauled skill tree might look like this:

(the XP and individual skill names are all placeholder p.s. sorry for the sloppiness!)
Every mech gets 20 skill points to distribute as they see fit. Unlocking six of the eight basic skills in any given level opens up the elite skills; getting three of the four elite skills unlocks mastery. This way, each mech can unlock the master-level skills for two classes, but only by going through a minial route for each, as well as locking out the third skill tree entirely; or else, the mech could specialize very deeply in one skill tree and unlock everything, while still having enough to get the basics in one other three, or match a little of both.
The skill tree I want to focus the most on is the recon tree, because it is the only one that does not have a precedent in MWO, but I will give some ideas for the others; once again, all my numbers and skill ideas are not set in stone, so give me your good ideas in the comments!
Recon, Basic:
1 Sensor range increase +75m
2 Can target 1 additional enemy (see BAP section above for mechanics)
3 TAG laser appears only in thermal vision
4 BAP though-object range +50m
5 Ability to call down 1 artillery/airstrike without a command console
6 All enemies targeted by you get a 10% missile lock speed boost
7 Allies within 100m gather target info 10% faster
8 ECM sensor denial range +25m
Recon, elite:
1 Command consoles can be used to call in artillery barrages using the battlegrid overlay
2 BAP gives mech info for one sub-target instead of just listing Unknown
3 ECM can create one fake IFF signature
4 Mech gains an additional consumable slot that can only be used for UAVs; 50% discount on UAVs for auto-refilling on this chassis
Recon, master:
1 Mech gains three additional module slots
2 Double all basic skills
These combinations of skills should provide an incentive to specialize in information warfare while also giving enough flexibility that not all skills are necessary (i.e. mechs with no energy slot can skip the TAG skill). Again, these ideas are just what's bouncing around inside my skull; if you can think of better ones, by all means post them, and I might add them to this post! And for those wondering, yes, the doubling of basic skills would no longer be automatic under this system, it becomes a master skill unlock for each tree. The idea is to create an actual difficult choice for the player between fully unlocking one tree with either an excellent skill or doubled basics while also mastering another tree, or letting a player fully specialize but locking out another tree's master skill tier.
The movement skill tree functions much like most of our current skills, and focuses on providing turning radius, and speed boosts, but also covers jumpjets and armor boosts now; the armor is split off the combat tree to make that one less of an easy decision. A possible consumable for a jumpjet fuel boost that works similarly to how coolshot currently works would add additional utility.
Combat is the damage-specialization tree. It encompasses the current coolrun/heat containment skills, as well as allowing the pilot to specialize in beam, ballistic, or missile weapons. It also gives combat targeting bonuses, including faster info gathering, and hopefully a magnetometer, which would act similarly to thermal vision but include short-range through wall visibility, but severely hampering the pilot's ability to see. The combat tree's consumable focus would work on making coolshots more cost-effective.
Because customization is a fact of life in MWO, this would also mean that skill trees would have to be refundable for a nominal c-bill fee.
The tl;dr? Information warfare needs a huge overhaul, and overhauling IW would be a huge step towards giving MWO real role warfare; overhauling both IW and the skill tree would add immensely to the game's depth.
Again, I'd love to hear what you think! Like it, hate it, think it has some decent ideas but needs work? Tell me! The best way to make a good ewar system will include feedback from the whole community.