'We're calling pricing a tie since neither GPU costs significantly less than the other. Of course, the AMD GPU was already declared the winner in performance, meaning it's a better value overall at the same price, but we already gave AMD credit for the performance category.'
I'm not one to go all fanboy over faceless corporations, but I can't understand the reasoning used in parts of this comparison. The 3050 was bad value from the moment it was released, and it's still bad value now.
The above part sums it up for me: why the aversion to giving credit to AMD when the performance of the 6600 overall is at times vastly better than the 3050 for a similar or even lower price? Because 'that wouldn't be fair to the 3050' is what it reads like to me. It makes no sense. Any extra 'features' or 'technology' the 3050 has can't make up for that level of discrepancy. You don't seem to have a similar problem with certain other product comparisons. Not to mention that the performance difference between the two cards may equate to going from a jittery mess on one to relatively smooth on the other.
I hope you're not using the overall geomean result as a justification for making price a tie - in general, raytracing performance is too low to be considered worthwhile for either of them. You make mention of this in the article, notably with the 3050 being 72% faster at 4k; a lot of that can be attributed to just Diablo 4, but it seems the performance margin is still being counted as valid. These are NOT 4K cards, even with rasterisation.
---
I've just read all that back to myself and realised I've gone off on a rant. And I don't care. I can't help it: I honestly think the 3050 is one of the worst value products Nvidia has ever spawned, and hate that so many people have blindly flocked to it. It's nonsense like that which enables companies to justify higher prices.