...do you have any evidence of this? I've calculated averages of games based on the scores listed on Metacritic for statistical purposes, and the average has always been the same as the Metacritic score.

I would say between 92 and 95. Just because Nintendo doesnt give Zelda that Hardware what it would deserve. Zelda deserves minimum 1080p 60fps. Such a beatifull game but the HW is holding the Devs back.

I am thinking about 97.5, no lower than 96.5, no higher than 99%.

They say it right on their website:

http://www.metacritic.com/about-metascores

From what I can tell, no they do not break it down specifically.

From my experience, I suspect averages are rarely, if ever, weighted. I've yet to find an example of an Metacritic score that did not line up with its standard average, and given that there is no list or explanation for how reviews are weighted, it seems more likely that this is a protection clause for potential future weighting rather than something that is done consistently.

Jesus im not saying its going to get a 98/99. I think personally a 95

Good and bad are relative terms. While i personally think 'bad' is excessive, relative to current AAA standards BotW's graphics are certainly sub-average.

To your second point, in the context of video games art style and graphics have always been treated as separate by many people. You're right that the end-result is tied to both, but they're still two separately defined properties. If you look at BotW and think it looks visually amazing, then you are in-fact separating them. You're saying it's sub-average tech (relative to market standards) doesn't get in the way of how you feel the overall game looks. If they were critically inseparable, then the lacking technology would drag down your impression of the overall image, but because you're able to see them as distinct elements with subjective impacts you're able to devalue one without affecting the other.

There are plenty of other games like that too. I think Persona 5 is one of the prettiest looking games ever made, but graphically it's decidedly sub-average.

Probably not. Not that it matters anyways. GTA 5 has a 90 something on MC and I hate the game.

They don't, but their weighting has little effect. The metascore is rarely more than 1 point off the proper average, so i'm not really sure why they bother. I assume it's there to mitigate the effect of extreme outliers (e.g. a game with 90+ getting a 1/10 from a small site), but we can't know for sure.