One of the things I used to do when working in large teams was wander around and look to see if anyone was having trouble debugging. Sometimes you'd find someone who'd been stuck on a bug for several days. So what I'd do was I'd sit down with them, and start putting in all the 'fluff' like indenting their code, splitting lines with multiple variable declarations each onto their own line etc. Essentially putting in all the 'long cuts' that they'd taken 'short cuts' around in order to 'speed up' the programming.
Often either they or I would see the problem immediately.
These days of course there are tools that do the auto-formatting, but I still try to optimise for readability, simplicity and elegance in my code.
Back to the lines of code as a metric issue (tm) in that scenario where it takes 2x as long to write a line in language X, and that line does twice as much, it is probably at least twice as complex as the 'more verbose' language... and I think this implies that it will be harder to debug and maintain later on. So arbitrarily complex languages are optimal for some internet 'toss off' (one off, toss away) competition, but not necessarily optimal for long term efforts.
Some of these languages that allow you to redefine the language itself on the fly allow you to do 'cool stuff' (tm) but tend to attract people who think they are much cleverer than they actually are. I think it is a regional variation, but all the Ruby fans in my city scare me. The thought of what would happen if I'd spent days or weeks on a bug only to find out it was caused by someone else 'tweaking' one of the core classes and not bothering to tell anyone else... twitch