The wine internal libraries are still about a megabyte of code, and it's silly to duplicate all of that whenever a single function changes. Symbol versioning is much better solution to that problem, and I don't see much reason for not using it (except on platforms that don't support it of course).
Yes, fair enough. I hadn't realised that while libwine is fairly small, libwine_unicode is huge (not sure what it does, but i expect i'll find out soon enough).
I don't really have any objection to symvers, other than it does mean it's harder to eliminate "dead" legacy code. That's only really a problem if the library changes fairly often though, and even then it's not such a problem these days.