mIRC does not handle CP1252 (or any codepages) anymore. This is an intentional design choice for Unicode compatibility. You should no longer be using any codepages in mIRC.
There is also no rule that says to be a Windows application it must handle CP1252. When your web browsers are set to UTF-8 content-types, they do not handle CP1252. Certainly web-browsers are Windows applications.
It should be noted that any CP1252 support in Windows applications is likely legacy only.
And no, the reason $utfencode differs is because you're attempting to $utfencode already encoded data; all data in mIRC has already been encoded in UTF-8, so calling $utfencode will double-encode the data. It would be equivalent to the following code in 6.35:
//echo -a $utfencode($utfencode(A))
You should never use $utf* functions in 7.x, full stop.
