I think they're calling it unfair due to power consumption difference.
Strix Halo in NUCs uses around ~180w when running heavy games (e.g. cyberpunk). Even in laptop form factor it still uses around ~100w. In contrast, the max these panther lake chips use is 65w.
The Arc B390 is not gonna be anywhere near as fast as a 8060s. It's not even going to be equal to an 8050s either but it doesn't need to be. If it can get close enough to it (and make up the remaining differences with it's superior upscaling) while being significantly more efficient, that's a win imo.
I think the B390 is going to be midway performance wise between 140V/890m and 8050s or about 50% faster than 140V/890m in pure raster when you limit to like the same low tdp (e.g. 20w). I know the 8050s is a lot faster when running at high tdp (e.g. 80 watts) but it'll be interesting to see how the power scales once you limit an 8050s to 20w. Or how close the B390 can get to it.
Actually, nvm - we already know, heh. :)
At 20w, the 8050s performs the same as a 890m/140V. So basically if you care about battery life while gaming, panther lake is gonna be 50% faster at same low power.
I actually really wanna know how the Adreno X2-90 in Snapdragon X2 Elite performs in comparison now. Even though, they were the most behind last gen, they seemed to have made the biggest gains. According to Qcomm, 2.3x improvement compared to their previous gen, X1 Elite or 50% faster than lunar lake. If true, that would make it almost as fast as panther lake..