Although an individual takeoff at full thrust will save fuel on that takeoff, the fuel burn will be higher over an average life as the deterioration in engine performance increases.
Quite so, when comparing e.g. 1000 full rated takeoffs to 1000 average reduced thrust takeoffs. No argument here, and I believe GE (or CFMI or R-R) have made similar statements.
The scenario I proposed is somewhat different: I'm comparing 10% thrust reduction to 20 or 25% reduction. The deterioration effects are quite non-linear; it's not evident to me that the additional performance retention of a higher flex temp will pay for the added fuel burn due to slower climbout.