If you make sure to run update.php, it's very rare for your wiki to break, unless you've hacked things or not updated your extensions or such. We're usually pretty careful to avoid significant regressions when upgrading wikis that are using supported/sane configurations.
Can we make update.php ask the user if he wants to install the new extension?
It's a regression for people who already have math working. What's the advantage? We have an awful lot of marginal features in core. When have we ever split a feature that we'd released in core into an extension, when a significant number of people were using it? I don't see the point. It's not like we're going to significantly reduce the size of the tarball or anything.
Is there any place we could get usage statistics for the math feature? I think the advantages for new installations justify inconveniencing some existing users, especially if we can automate installation of the new extension, but this discussion would be better with some data.
Yours, Damon Wang