Not saying that LLM development isn't fast, but doing a re-implementation of something you know in-and-out is almost always going to be faster, as you skip the probably most time intensive step of product research and requirement engineering.

Absolutely, but that doesn’t discount that it’s now feasible to spec a feature once and have it built for me on both apps. I’m doing the same process on a new project I’m working on and it’s built both apps simultaneously as I’d hoped.