And while you are at it, producing show off cases, you will find that you will learn a broad set of areas if you are anything like me. Say you start out to so a simple web page, with some dynamics. So you check ot CGI, and get the hang of it, pretty soon you will want to test this locally. So you install, say apache. Then you want to use some SSI, or mod_perl, and you want to password protect a directory. So you learn som basics on how to configure your web server. Then you decide you want to have users that can log in. So you start to learn DBI, SQL and the tweaks and quirks of a database. Not to mention, you learn how to install and administrate these things. Then you are concerned that someone might hack your users - so you read up on securiry, encryption, maybe even play with SSL.
You see where I'm getting at, yes? Something I think you should do, and this is a part I'm having trouble with privately, is to find *actual* projects you want to do. It might be something that scratches and itch, or just something you find interesting, whether it is your own site about your hobby, or a word processor for sci-fi authors.
Whatever your choice is, you will learn lots, and automatically broaden your view.
If you want to be good at perl, don't think too much about what parts of you need to know, just try to be a fluent speaker of the language, and learn how to find the information and the modules you need. This is one of the cornerstones of productivity in my world; the ability to find the tools you need, and the manual on how to use them when you do need them.
Hope that helps. :)
You have moved into a dark place.
It is pitch black. You are likely to be eaten by a grue.