Probably one-third of the projects I have worked with did use a home-grown solution, especially when those projects have been around for a long time such that they were originally built for much-smaller hardware.   I suspect that many of the more-lightweight tools that we see today in CPAN are contributions made from these.

If you are at the luxurious point where you still can make a decision, knowing of course that whatever decision you make will wind up being permanent, I would strongly advise anyone to try to make the decision as a team, by consensus.   Have several team members choose a different CPAN tool, and have all of them take a couple days familiarizing themselves with the tool and writing some moderately-representative web page using it.   Then, circle the wagons and have the team members de-brief the others.   Invite each of them to present at least three reasons why the project should use this tool, and at least three other reasons why not.   Remind them that all reasons should be technical.

I would suggest that efficiency – that is to say, machine efficiency – is probably a much-less important factor today than it was in years past.   But, team efficiency, maintainability, and expressiveness are more-so.   Whatever tool you they choose, realize that the choice is being made forevermore.

Indiana Jones and the Last Crusade:   “He did not choose ... wisely ...”