It seems disproven, or perhaps only useful to some subset of people they haven’t pinpointed.

https://www.sciencedirect.com/science/article/pii/S0001691824001811

Honestly though if it helps someone they should ride that placebo as far as it takes them!

I read the study @BCOVertigo@lemmy.world linked and the following statements can be made:

for 32 university students with normal vision and without dyslexia who have not encountered bionic reading before

• There was no statistical relevant difference between bionic and normal texts regarding reading speed and comprehension
• There was no statistical relevant difference between faster and slower readers
• There was no statistical relevant difference when given time to adapt to bionic reading

but

• in bayesian analysis there was no strong evidence for either the null- or the alternative hypothesis (meaning an effect cannot be ruled out) even tho the sample size should have been enough to do just that, meaning it could still have an effect, but it would also mean that if it did, it wouldn’t be something to recommend to someone out of the blue because the relevance for singular people is negligible
• the study did not test people with dyslexia, ADHD or vision impairment, so for those groups no statement can be made.

There were no studies following up on these results, so if someone wants to go for it, it’s free real estate!