So I'm wondering several things here... First, who's paying for this survey? I note that many surveys produce results pleasing to whoever is paying for them. Second, as was already mentioned, what are the real statistics? Without some idea of the sample, you have no idea how skewed the survey might be.

There is a small book called How to Lie With Statistics by Darrell Huff, that is most informative. He discusses many common mistakes - or outright fraud - that is presented as fact. Without the supporting information, there is no way to know how valid this survey is.

I think that Perl, at any rate, has a bright future (of course, I'm not objective...). It is extremely useful, has an enormous archive of modules, has a long history of use, and a new version is in the works that will likely address many of the perceived shortcomings it is said to have. Moreover, the number of programmers who know Perl must be enormous. They'll continue to use it, as long as it's available (unless by some unbelievable means, something better comes along - won't happen for a long, long time, I think).