Now that’s a most peculiar “paraphrase,” good sir!   And you yourself know it to be a lot of bilge, so there.   (You’ve got plenty of industry experience.)

Software takes a fairly huge amount of money to get right, but it costs next-to-nothing thereafter to keep it that way.   Sometimes people make the serious error of rewriting an application that works, in a “new and improved” language, only to discover ... the hard way ... that three things were true:

1. They didn’t actually know all of the things that the original application did, nor exactly how it did them.
2. That they seriously under-scoped and under-estimated the project, and failed to identify all of its subtle interactions with other parts of the business.
3. That, when the du$$t finally $ettled, what they’d come up really wasn’t “–er” than the system it replaced ... or, as the case may be, failed to replace.

If you have a system in Perl-5 that right now is doing exactly what you need it to do, there is little cost and little business risk in keeping it that way.   Whereas, if you set out to replace it or even to move it to a different language platform (or version), both the costs and the risks go sky-high.   You’d better have a rock-solid business case for that.

These “popularity contest” polls merely explore what programmers say they would prefer to use, when embarking on a new project, all things being equal.   As such, they’re really not worth much at all.   Call it “maintaining what you’ve got” if you want to, and decry it if you want to, but that is most of what any production shop actually does.   As you know.