Maybe this comment is already here and I just missed it, but languages naturally have life cycles, and thinking in terms of Ruby's life cycle might explain both some of your early enthusiasm and your current discontent. (I'm going to draw in broad strokes here and will probably get some details wrong.)
In 2002 Ruby was young: half-formed and exciting. It's problems didn't really look like problems: they looked like things Matz hadn't gotten to yet (and, even better, opportunities for other programmers to get to first). In late 2008 Ruby is established; the problems come with a set of reasons why they aren't likely to be solved. (The other problems were solved.)
This is natural and normal for any multi-year group endeavor, be it a company or a rock band: it starts with a world of potential. If it is one of the lucky few to survive for any length of time, it becomes mostly a specific reality, with a humdrum future. (More often it goes from having a world of potential to not having any future at all in short order.)
You cite as counter-examples Javascript and Scheme. Javascript is benefitting from the HUGE externality of being the client-side language for the single most important application category going. As for Scheme ... there's a paraphrase of "Brazil is the country of the future, and always will be" that's escaping me, but suffice to say that if the 73d time is the charm for Lisp's taking over the world, I won't be too embarrassed to have bet against it.
Ruby's momentum 2002-2007 was like Python's momentum 1999-2004 was like Perl's momentum 1993-1998: the momentum of a firework or a homerun. It was glorious to watch them rise; it shouldn't be sad to see them return to earth.