To a ridiculous extent, they're all just languages. You can write good programs, and bad programs, in any of them.
So, in the end, it comes down to how fast / cheaply I can deliver the results a customer is looking for. If the customer already has an array of applications he has to look after in technology A, then using technology A for his problem has to be top of the list.
Well, that's the way I used to think, and I've been doing this for well over thirty years now. In assembler(s) of various ilks, basic(s) likewise, cobol, DL/I, Erlang (nah, just threw that in for the sequence), Fortran... Most recently in Java, Oracle, Perl,
During this year, though, I've stumbled on XQuery and the eXist database. An XQuery can deliver XML, XHTML, or any other variant that's needed. If it's been invoked from a browser it delivers it back to the browser. So, all I need to deliver interactive functionality on a given set of data is a single XQuery (text) document. The change, test and try again cycle comes down to about 60 seconds or less, as all I do is change the text document, save it in the repository, and retry the browser request. If it needs more detailed testing (and, since I'm learning this stuff as I go, it often does) I can drop the source into the eXist XQuery sandbox and work it out line by line.
Now I find myself trying to explain to an organisation with twenty years of dedicated Oracle tradition that it is in fact Oracle that's causing their long decline in agility. Worse, the patent battles playing out before the US courts have stripped the net of uptodate information that would support my cause.