Well, on my system at least the GUI of XChat is very slow, and entering text with the Japanese IME doesn't work properly aswell as it only takes the first character of the entered string. And besides, saying it sucks is still a personal opinion

.
About UTF-8 as a default encoding... well... yeah... uh... well... in the end it's best ofcourse, but all I care about at the moment is that I can properly read and type UTF-8 ^_^. I think it should be fairly easy to make it a per-channel property, as the font can also be set per-channel. And if the UTF-8 encoding is set, convert all incoming channel text to whatever internal format mIRC is using. Gah, at least in C# and with .NET this is all a non-issue as everything uses Unicode and there are easy converters available ^_^;;. And the NT-based windows offers plenty of unicode functions in its 'regular' API.
About joining several channels and servers... Scripting mIRC is pretty easy... I believe /server -m <server addy> <list of channels> does the trick already (from memory), and make that execute as mIRC starts, or something.
By the way, as I read a little further into UTF-8, there seems to be some 'nastiness' around this topic. Well, at least people aren't always friendly to eachother. I hope people won't flame me for making this request and tell me to use the search function bla bla... I should have tried that one before posting, but then again, would you rather have me pulling an age-old thread to the top again by replying to it?
~Grauw