This is exactly how I have always worked. I personally don't operate well breaking things up into small, bite sized PRs like people want. So instead I do the whole big project, and then break it apart into logical segments for my colleagues.

And this is often how I proceed when working with an LLM as well. We build the whole thing, and if I think it needs to be split apart, then in another session I work with the LLM to break it apart into reasonable chunks.