Prices rising with inflation is only keeping them stable in constant dollars. You'd have to apply an inflation discount to your future revenues as well.

My model isn't trying to argue that you'd value the actual real-world value of a book in terms of a constant, real-value, revenue stream. Reality is far less remunerative than that.

It's that even if that were an accurate model, the distant-future revenues would provide a minimal incentive to a current opportunity-cost analysis of writing for publication and income vs. alternative activities.

Which is to say: copyright terms are far longer than can be justified on an economic basis as incentive to authors.

That they're an incentive to present copyrightholders of aged works with some present value isn't at debate, but that's the only justification for post-hoc copyright term extension.