> maybe it is just more comfortable to read, not necessarily faster?

The article says that participants in the studies preferred the traditional fonts over the dyslexia fonts. I would argue that this contradicts the thesis that they would be more comfortable to read. Moreover, the way I read the article, it wasn't just reading speed but accuracy that was tested as well.

> There are more and more cases where my personal experience seems to contradict with science. And I am not sure what to make of that.

I find that I often have to question my preconceptions when I encounter this issue. In other words, I have invested e.g. time, effort and thought into something which I thought works and it is difficult to not fall into a kind of sunken cost fallacy, i.e. my brain doesn't want me to believe it does not work, because I have invested effort into it.

I only looked at the study about the open/free font. Two things I noticed were that the experiment design seemed to use lists of words, instead of reading in sentences or paragraphs, and that the base performance of (for example the correct letter rate) the traditional fonts were also very high (basically already 100%).

It's possible that the test used does not generalize to other reading contexts and populations.

I expect familiarity would be a big confounder either way? The first time you see the new font it might be harder than something like times new Roman if you've seen that thousands of times and gotten used to its hinting