platyhsu One reason can be that Dan ran the experiment in Novemeber as per his post. Since then we had a lot of changes as Kagi is constantly improving based on feedback we get from users.
But I think more importantly I think the main problem with the premise of the article is that the author's conclusion does not work for any user type. if marginalia is the best search engine out there as the article implies (I wonder if the author indeed uses it as default), it would have much more people using it. Likelier scenario is the choice of queries and affinity of the author made it look that way. The question of what makes a good search for more general population is much more nuanced.
Ublock is no.4 for me for "ad blocker" on Kagi, higher that in any other search engine. While it would be ideal that it is no.1 what I disagree with is the premise of the article that Kagi results are bad or terrible. For me (may be biased) they range from great to good on author's sample of queries.
Not saying there isn't room for improvement, but one thing to know is that a general search engine has to be designed to be used by everyone, not just tech-savvy crowd. My wife would probably prefer the vancuver forecast page that is no.1 in kagi than official ca gov meteo site where I have to spend considerable time to summarize the answer to my question from the data avialable.
Being a general search engine makes it a hard design challenge and this is why we built a number of tools that no other search engines has, to let users personalize their results further, because no two humans need the same search experience.
