I tweaked the list of unit tests a little more, trying to incorporate some from your Docker build settings.  Not sure what's going on with the pickle results.  Below are new pyperformance runs, comparing my PR to the "master" version it is based on.

I added the --with-profile-task configure option.  If someone really wants to wait for a long build to gain a tiny bit of extra speed, that makes it pretty easy to do so.

BTW, I tried making the --pgo disable most of the gc_collect() calls.  It doesn't actually make much difference in profile generation time so I discarded those changes.