I believe that most hunting related things are used in loot calculation. But I think there is a deminishing return for inefecient decay. IE the first few uses of your fap count the full decay in loot but the more you use your fap on a single mob the less is counted for loot.
DISCLAIMER: This is a personal theory and not something I am stating as fact. Also feel free to check my math as it's not my strong point.
Based on my observations I believe it works something like this...
Two factors influence the amount of decay being added to your loot.
1. FAPs have an invisible "efficiency rating" similar to weapons which directly impacts the maximum % of decay returned upon use. Call this variable A.
2. FAP decay is added to loot based on the % of actual heal used. Call this variable B.
I'll use D to indicate decay in this equation.
So decay return would look like this... (D x A) X B = Base Return
So lets create a hypothetical scenario...
You have a FAP with a max efficiency factor of 95%. This FAP decays 10 PEC and heals 100 HP.
If you are missing 100 health, then it looks like this....
(10 * 0.95) X 1 = 9.5 PEC added to loot.
If you are missing only 70 health, then it looks like this....
(10 * 0.95) X 0.7 = 6.65 PEC added to loot.
Conclusion: Over-fapping results in PED lost (presumably going to the bonus shrapnel pool) while maximizing effective healing results in maximum decay returned to loot.
A few notes;
1. There may also be a third factor, something like a threshold factor allowing for a less than full heal to be used while still returning the max amount of decay into loot. Consequently some faps seem to be more "forgiving" of overhealing than others.
2. I do not have substantial empirical evidence to back this theory up. I conducted a handful of tests of various faps using turrets, drowning, and shared loot mobs and "eyeballed" my results. Because I was doing this exclusively for my own information I did not meticulously document my data.
3. The theory of the "efficiency rating" factor is purely speculative, and would be much more difficult to test than the actual heal factor.