I’m generally anti- modules. I’m also iffy on variable end game scoring. Both of these, suggest to me, a kind of designer that hasn’t quite grasped what they’ve made and is struggling to pin how best to make that design shine the best.
I think though there are some interesting caveats to this thought. For one I think there’s a kind of Knizia thing where he makes like 10 games which have the feel of one game each but with a different variant. But he quite often sells each of those as a singular product. I think this is obviously more expensive but it does create a smoother product (eg a single rule book that knows what you have, a board which doesn’t have bits jutting out here and there)