Bearing in mind that IANAL:
Making a copy of the music for your own use
is fine, although the RIAA will tell you otherwise. The DMCA doesn't apply because you're not defeating any copy protection in ripping a standard CD.
But, as Darwin noted, sharing the results with others is (depending on where you are) almost certainly illegal, regardless of whether you use web sites, FTP, or what-have-you. In practice, if you put the music onto a server that no one knows about, with no public links to it, and they're stored in password-protected archives so that no one outside your family can actually see the content, then you'd probably get away with it. But I'm not recommending that.
I also agree on the metadata. AFAIK, all the various tools (WMP, iTunes, or my favorite ripper, Exact Audio Copy) rely on the same CDDB service. If WMP's query to CDDB doesn't yield results, then the same query made by EAC won't fare any better.
For commercially-produced CDs of contemporary genres this almost never happens. But it's different for classical music. For one, classical music has a smaller audience than NSync (sad as that might be). So there are fewer people entering the data.
But also, classical CDs are frequently "shovelware". Companies can produce these for next to nothing (no copyright royalties, just pay an orchestra for its time and you're done), and thus reap huge profit margins. I'd guess that there are hundreds of versions of, say, any of Beethoven's major symphonies available for purchase. The identifying characteristics of one bargain-bin copy of Beethoven's 9th are completely unrelated to a different recording of it.
The result is that any given classical CD has a much lower chance of having a music lover enter the data into CDDB.