Perhaps you read, or skimmed, the story linked off of Slashdot about the module count for Ruby growing to eclipse the module count for Perl? First things first: it's wrong.

Second: is measuring module count useful? In some respects it is. A community with 10 modules would seem to be less active than a community with 1000 modules. But given the number of modules in each of the 'big' communities (Perl, Ruby, Python) it seems less useful as a metric. So what would be a good metric? Module quality? Module popularity? How can those be measured?