Ruby, JS, Perl, etc. can't really touch this. There are all sorts of "flavor of the moment" languages right now on HN but at the end of the day, if you want a mature, workhorse language that's easy to learn, has a lot of good libraries, and has a great community around it, choose Python.
Most modern dynamic programming languages owe a debt to Perl, and the modern web certainly does. It has a variety of modern frameworks:
I'd say Catalyst and Mojolicious are the one's I've heard about most.
Perl may not be considered as cool as Python by some, and there are reasons you can criticize it, but it's not any of the reasons you gave.
By "flavor of the moment", I was not referring to Perl, but rather to all the "compiles-to-JS" type stuff.
Merely comparing package repository sizes is not any indicator of quality or future prospects. Once you're beyond a few tends of thousands of packages, the bases are basically covered.
The biggest ding against Perl is the readability and maintainability. For scientific and data analytics oriented code, generally written by non-programmers who do not understand good software development process, Perl is frequently way too much rope.
Coolness has nothing to do with Python's popularity. (Unlike e.g. Rails which soared to popularity as a "Non-Java" way to do web apps, and then became a bit of a victim of its own success.) The scientific Python people I know all moved to Python from Matlab, C, and Perl because it made them more productive.