Originally Posted by HTupolev

Okay, so your point was actually just that the power is too low relative to the heat dissipation ability of the frame to make a noticeable difference.

Originally Posted by HTupolev

This is obviously true in the real-world case, although whether it would remain true in a hypothetical case where internal frame friction caused significant power loss depends on what we'd consider "significant." Note that one person you've been agreeing with in this thread - raria - has previously pondered whether flex about the cranks was slowing them down by three miles per hour, and has seemingly even purchased a new crankset in an effort to address the issue.

I agree that, with the frame being metal, there's probably a large range of hypothetical dissipative power loss in the frame which would be considered a meaningful performance loss to many people, while not being significant enough to cause frame heating that would be very apparent to a person.