I suppose it depends on what you want out of it. To create a complete experience (the main goal of 7drl I think)? To conduct an experiment? To build a platform to build on?
If it's the first one, undershooting and polishing instead seems correct.
It's correct in the sense that you do end up with a result that meets the goal. It also leaves potential on the table that could have gone toward "more game."
If there are two hypothetical undershot projects that both end up polished and complete but differed in level of ambition, I suspect we'd agree that the more ambitious one got the better result.