But there's a high correlation between companies which trust their programmers to use the right tool for the job and companies which "get it" (it being how to develop and maintain technology).

Will that eliminate 99/100 companies? Probably not that many but it will eliminate a lot, and you should be glad.

After my first job, I learned to always ask what a company is doing for source and configuration control. If they have no answer or a bad one (Visual Source Safe), that will tell you a lot about the environment.

Same thing if it's an MS shop that refuses F# or a java shop which refuses Clojure a priori.

"Getting it" may be right for us (developers) but wrong for the company. A florist that needs order tracking isn't trying to change the software world. That persuasive guy thats insists SmallTalk is gods gift to florists is NOT doing them a favor.

Perhaps not, but writing it in Java isn't exactly doing them a favor either as it increases time and cost for development and changes. The Right Thing for such a situation is probably a compromise between language power and popularity. Ruby or Python with one of the popular frameworks that makes half the decisions for you would probably be a good choice.

I've replied to this at length:

http://news.ycombinator.com/item?id=901411

Estimates have it that there are about 50k people being paid to write Ruby on Rails right now, chances are that it's way higher because it starts to be the language of choice in outsource India because of the high conceptional similarity between different projects written in rails and how they help new teams get started with an existing code base.

I sure as heck can't find enough people to hire for ruby ( and even objective-j !!) jobs.

I'm not saying your figure is wrong but I think that if you actively look for a job in these languages then it's pretty simple to find one.