Maybe you're too used to the MS theory of bugs, where they believe it is better to deny the bugs exist rather than actually fix them
If you mean the Micro$oft that fixes more bugs and more quickly than any other software vendor known to modern man then perhaps you are right.
As for testing your urban myth that the case bug is the fault of an IRCd rather than mIRC, I cannot help there much because I only know of two networks that use CR (as I don't hop around as much as I used to) they being Bigpond (which doesn't currently allow user-created channels) and Webnet, where people who try to prove a point for the sake of trying to be right when they are wrong are too lazy to register nicknames. As for the second IRCd which affects mIRC this way, I do not know where that is. One thing is for sure, as I have stated, mIRC seems to object to being force-fed a changed case for inappropriately used modes therefore there is a bug in mIRC, as the original poster suggested.
