I tried to read carefully so please forgive me if I missed it, but I didn't see anything addressing response provenance or performance as in accuracy/other_metr

I tried to read carefully so please forgive me if I missed it, but I didn't see anything addressing response provenance or performance as in accuracy/other_metric. The difference between a web search and text generated by an LLM can be that a web search points to actually existing companies or entities that make statements. Sometimes this is required and summaries or paraphrase are insufficient. For performance or whatever metric one chooses, it's likely that web search wins some and loses some. It would be interesting to see an in-depth shoot out.