That *is* concerning. Thanks for cluing me into your computer platform. Apple ARM or Intel? The Apple version is built with a different compiler than the Windows & Linux versions--in such cases there is always an edge possibility that the significantly different environments would handle some bad code different than the other.
Unfortunately, I'm doubly boned in regards to reproducing & testing (and thus fixing) this issue myself, as at the moment and for the foreseeable future do not have access to my Intel Mac I use to build the Mac version. And I don't own an Apple ARM Mac to begin with. But I'm not without professional experience investigating this sort of hard to reach issues--it happens and these sort of the issues are the worst! I'll do what I can for the time being. But even if I suspect I found the problem and propose a fix, I won't be able to build a patch without my Mac.
As for making the last round auction skip an option--that is definitely a possibility and I will definitely promote it as such when possible and enough people complain about the current behavior.
Thanks for the feedback, Dan. If ever I get a test build for the mountain slowdown issue you mention, I will likely seek your help in testing / verifying an intermediary build.