There are absolutely valid questions around how much diminishing that compensation affects innovation, and around the marginal value of each year of copyright validity.

The way I envision it, there is a baseline of innovation. I.e. innovation that would occur with or without copyright. For each year added to copyright length, I assume that some amount of compensation to the copyrighter is added (although the marginal compensation likely follows a bell curve). There is also some degree of cost, as other people are unable to innovate on top of those copyrighted works.

I think we can both agree, there are people who innovate because of the potential compensation. There are also people who will not innovate because of the lack of compensation. The lack of investment and innovation in antibiotic medicines is a good example of this.

The question is the optimal balance between compensating people who innovate and allowing other innovaters to build on prior work.

Academics have already calculated the optimal balance, in particular the optimal copyright term. It's probably in the region of the original 14 years instead of the current ~150. https://rufuspollock.com/papers/optimal_copyright_term.pdf