Because light and medium mechs are not as popular to to play, queue times are frequently longer than they should be. A little extra encouragement to play the less popular stuff could help that quite a bit.
My dim recollection is that the supposed lore-typical breakdown of percentages of mechs actually seen in combat was something like 40% medium, 30% heavy, 20% light, and 10% assault.
Based on where you typically see mechs placed, lights are usually only in Alpha lance, mediums in alpha or bravo, heavies in bravo or charlie, and assaults in charlie. From that, we could assign values of 17/33/33/17 to light/medium/heavy/assault as ideal.
We don't usually see anything close to that. (frequently: heavies ~40-50%, assaults ~20-30%, light ~10%, and mediums ~20%)
However you come up with the ideal numbers, I would actually tie the rewards in to the actual percentages of mechs being played, updated periodically. I would hope that PGI has the data mining capability to do this. By tracking what the actual numbers are (factored hourly?), then comparing them to what the ideal is considered to be, you could apply a bonus based on how far off the numbers are. I would NOT punish folks for playing the popular classes.
PGI has already tried to encourage the use of lighter classes by improving the xp/cb rewards for non-kill/damage activities. That's great and all, but a) it doesn't seem to be well reflected in match score, and

More needs to be done.