I've found a problem with Texinfo 6.6 which happens only when using TEXINFO_XS_PARSER=1: any UTF-8 encoded text in the Texinfo sources is corrupted in the output. For example, the UTF-8 sequence \303\240, which is the encoding of à, becomes \303\203\302\240, and \342\200\230, which is the encoding of ‘ (left single curved quote) becomes \303\242\302\200\302\230. Try generating elisp.info from the latest master branch of the Emacs Git repository to see this.
If you don't see the problem on your system, try doing the above in a non-UTF-8 locale (e.g., a Latin-1 locale). If that doesn't succeed in reproducing the problem, either, it could be Windows specific, but in that case I will need your guidance for where to look. Thanks.
