A high-quality replica takes tremendous effort and time, but it’s not really financially viable.
I’d be happy to be proven wrong.
Oh, and I’d be super happy to pay for such a replica too! For example, $99 a year for a Mac OS X Jaguar to Tiger era skin/theme/desktop environment.
KDE, GNOME, etc spend a lot of resources on new features and overhauls to enable new features. A fixed-feature replica project by comparison would eventually achieve a “finished” state, with resources instead going towards optimizations, bug fixes, code correctness, etc which are overall less demanding and make unavoidable tech migrations (e.g. X11 → Wayland) less of a pain since the project won’t be mired in tech debt.