Starting from scratch usually sounds like a great idea. You don't have to deal with legacies and workarounds, what's not to like about that? :) So I get it why people think it's best to start GIMP anew. I also get it why people are reluctant to do so. While you are working on something new, you also typically don't have a working program for years. And when they already have software that works, that means maintaining two code bases at once.

In case of GIMP, the existing team is already stretched thin and has to deal with too many reports and requests to handle. In the past 4 years, the amount of reports in the issue tracker more than doubled, but the team hasn't grown accordingly. I don't think it's realistic to expect that the team would be able to maintain the current code base while starting anew. If the current progress is already commonly referred to as glacial, what do you think will happen if they start reimplementing GIMP from scratch?

And I get it why you are referring to Photopea here. I think Photopea is a formidable effort. But it's been under development for 10 years already. That's quite a long time. And Ivan didn't have another image editor to maintain while working on that project. AFAIK, he originally intended the write a web app to merely open and show a PSD file in the browser. That's a quite different development trajectory.