You are overstating the results of such studies or use your datapoint out of context. Just because A is on average better than B, does not mean that A is better for everyone.
If you have a study that shows that most people do(n't) like pineapple on pizza (maybe make a blog post "is pineapple on pizza such a good idea?"), that is rather unrelated to someone likeing pineapple on pizza or not and is not claiming that all people don't like pineapple on pizza.
Darkmode is a nice thing! It feels like some people hype it as the 'one true better solution to everything' or a strictly superior version, when it boils down to preferance and minor differences. In the end you probably want to have both options (if you have the capacity to support them).
If you have a study that shows that most people do(n't) like pineapple on pizza (maybe make a blog post "is pineapple on pizza such a good idea?"), that is rather unrelated to someone likeing pineapple on pizza or not and is not claiming that all people don't like pineapple on pizza.
Darkmode is a nice thing! It feels like some people hype it as the 'one true better solution to everything' or a strictly superior version, when it boils down to preferance and minor differences. In the end you probably want to have both options (if you have the capacity to support them).