|There's more than one way to do things|
As an initiate I'm probably in the position to have a go at this question from the 'outside'. So please bear in mind I don't want to offend anyone, but only want to give my point of view on development of 'large' webapplications. It's my intent to add my 5 cents to this discussion (where possible).
I don't know (yet) if Perl would be the proper language for -your- project to build in. In my opinion there are more perspectives to take into consideration for that. I will name a few; the architecture it is suppost to fit in, the duties it has to perform (interaction), the platform it has to operate on, etc. Answers to those question should, in my opinion, also be taken into consideration when picking the language.
As for the age of Perl concerns, I would point out that this can also be an advantage. Bleeding-edge languages/tools are not always the best things happening to enterprise software. So when picking a language I would also take into account the expected lifetime of the application and the fact if there are significant other reasons (as compliancy, etc) which have to be met.
As for frameworks, I think there isn't a framework around (in any language) which hasn't have a learning curve. It is worth the time to investigate what this curve might be, but also what the pro-and-cons are of some frameworks, starting with the ones in the language(s) you already have knowledge of and compare them to each other and to the requirements the application (and its environment) has to deal with.
After taking those items (also) in consideration, it might turn out that Programming in Perl of this website indead is the best choice (with or without a framework), or not.
But as always ... since there are always more ways of doing things, there is not such thing as the one-right solution.
small, perhaps trivial update. The question as asked is also a question I (still) have. So the answers of others are also a contribution to my learning curve. tnx.