What's the point? It doesn't matter what model he says, he'll either get "that model is bad" or "you're holding it wrong". I've used 5.6 Sol on 1M context and

What's the point? It doesn't matter what model he says, he'll either get "that model is bad" or "you're holding it wrong". I've used 5.6 Sol on 1M context and max thinking, and also Opus 5 on max context and thinking, and I agree with that comment. Now what? Does this information matter or will you just say we must be doing something wrong?