I'm going to take the path of more compression resistant material in the belly. Starting with thicker glass.
I once asked an experienced FG bowyer about using thinner glass on the back in order to raise the stresses on the back and use more of its capacity (FG bows normally don't use anywhere near the full capacity of the glass). He warned me against going too far with that as the wood core tends to fail in shear right at the glue line if you work the FG lams too hard. The problem is the FG can withstand strains that are 2-2.5x higher than the wood can.
Using thicker belly glass is OK, but you can't go too far with making the back lam thinner when building an asymmetrical limb stack.
One approach to knowing what is happening is looking for deflex (string follow). My bows are primarily ASLs which makes string follow easier to observe and more likely due to the higher ratio of core lams.
How much set do you see happen with your bows? A thicker FG belly lam might be a very good idea if the core is taking set that is measurable. It would shift the neutral axis and lower the strains on the core wood at the glue line to the belly lam.
What if there is no follow observed over time? Then an indicator would be the changes in draw weight over time. It calls for careful record keeping. If I can get to no string follow and no, or very little, loss of draw weight I'll feel that balance has been reached.
I agree, set shows up first in loss of draw weight (when tillering a wood bow, which is all my experience). That seems to be a very good way to detect it before it becomes visible.
I understand what you mean about being balanced now. You are trying for a no set bow in the end, which is a very worthwhile target.
Maple (if similar to other hardwoods) is said to be twice as strong in tension than in compression. whether it can stretch as much as the glass above is doubtful,
I use maple boards for bows and it definitely is twice as strong in tension as compression. No wood can survive the strains that FG can.
when sinew is applied to the back of a bow, it is sometimes mounded up so the outermost layer can stretch even more. Perhaps you could consider applying the extra glass you propose for the belly in a similar fashion, rather than spreading it across the full width.
Since FG is heavy this would probably not be the most efficient way to use it. I suspect you would get the same effect with a thinner lam that runs the full width of the belly and has less weight overall. Would have to crunch numbers to see how it all works out, though.
Mark