Yeah it makes no sense, other than if they were to argue that OOT's score is inflated because of the factors I mentioned - in which case I guess people can retort by saying the user score is higher so it's still better than GTA4 even when both have an 98.
Yeah it's sitting at 99 and even has a Guinness World Record for it, but like I mentioned above, someone who looked into it found out that the reason it's sitting at a 99 is because Metacritic inflated its score.
Metacritic did so by counting non-point based reviews and non-10/10 reviews as a 100, including retrospective reviews from way past the launch window, and leaving out reviews from certain magazines that were included on other games from that time period which scored it lower than 10/10. When you account for those things the score would drop to a 97.65 instead of 99. Which would be rounded up to 98 which is the same score that SC, GTA4 and Galaxy have (the same person also applied the same recalculation to Soul Calibur but it still maintained its 98 score regardless - and Galaxy/GTA4 were modern releases with plenty of reviews where this MC calculation is not an issue).
The main takeaway being that games from before Metacritic's launch (2001) don't really have accurate/equal review score weighings. OOT is probably not the only game affected by this but obviously it's the most prominent one.
But tbh even if the score dropped to 98, if you look at the pre-rounding scores of each of these games OOT might still be the number one (just like how Soul Calibur and Mario Odyssey are still ranked above GTA4 despite them all sitting at 98, and I checked because initially I assumed it might've been sorted by release date but for example Tony Hawk Pro Skater 3 has a 97 just like BOTW while releasing 16 years earlier, but it's ranked lower than BOTW).