There are these plugins:

https://github.com/zatevakhin/obsidian-local-llm

https://github.com/hinterdupfinger/obsidian-ollama

Which already exist and if nothing else are decent starting points.

> Not sure if local LLMs are powerful enough yet to enable meaningful/reliable outcomes

I've dabbled, briefly, with Ollama running Mistral locally on an M1 MacBook Pro with 32GB of unified memory, and throwing a couple of hundred markdown documents at it via a RAG resulted in quite decent output to prompts asking questions about abstract contents/summariesbbased on those docs.

So I'd say we're already at a point where you can have meaningful outcomes; reliability is a whole other issue though.