What is most apparent to me (putting in existing code and asking for changes) is Gemini 2.5 Pro's tendency to refuse to actually type out subroutines and routinely replace them with either stubs or comments that say, "put the subroutines back here". It makes it so even if Gemini results are good they're still broken and require lots of manual work/thinking to get the subroutines back into the code and hooked up properly.
With a 1 million token context you'd think they'd let the LLM actually use it but all the tricks to save token count just make it... not useful.