The first one was good, it just had a few glaring flaws that tarnished the experience; namely the most frustratingly boring boss battle ever.

It is indeed, in fact I'm writing this post on the beta version.

No, the point highlighted in the above article is that instead of making a case for specific types of fair use that might be applicable, the defendants lawyers tried to argue rather broadly that all downloading was fair use, something that the judge couldn't possibly find in favour of. In essence the judge said that they were open to certain arguments being made, but the defendant never tried to make them and instead opted for a bat-shit crazy defense coupled with irresponsible and outright illegal conduct. It's not hard to see why Tenenbaum wound up getting the judgement that he did.

I would suggest that if a web developer is incapable of making browsers render the same site consistently using CSS then they have a great deal more to learn. The problems with cross browser rendering issues have long since been solved.

You might also try

http://reference.sitepoint.com/css

It's an interesting idea, but like a lot of CAPTCHA alternatives it would only survive on a small scale.

The reality is it'd be quite trivial for a bot to request the page on which the form is displayed and retrieve the appropriate cookie and then subsequently parse the response to determine which form fields match up with which labels, something which is trivially done in any standards compliant website. At that point all they need to do is post to the server using that cookie and those form fields and they're through.

It's not so much making a parser for this syntax that's the issue, the issue is the ambiguity between functions in namespaces and static classes. See my post for further details.