As far as I know , MOND is not compatible with some observations. So dark matter and MOND have to be tweaked to agree with what we see.
That is a very, very fringe alt-science view that so far exists in various pop-sci blogs and informal youtube discussions, rather than in mainstream conferences, journals, and research centers.
I'm not saying it's bad to speculate or write whimsical blog articles about these topics, there's lots of room for creative speculation, but MOND does not respect Lorentz invariance which is basically a deal breaker for virtually all serious researchers. Fewer things have more robust empirical support than Lorentz invariance, so discarding it forces you to say that your violation is always just over the horizon of what is testable (and this horizon keeps getting pushed back). That's generally a discrediting feature of any theory and a huge red flag. Much better to say "you don't know" than to postulate something which requires such massive fine tuning to always be just beyond the horizon of what is testable, but close enough to that horizon to have explanatory power for observed phenomena.
https://phys.org/news/2017-11-evidence-violation-lorentz-inv...
https://asfriedman.physics.ucsd.edu/Papers/Friedman+2020a.pd...
But again, the 2017 experimental data [cited previously] in support of Lorentz invariance and the 2016 bullet cluster data in support of dark matter basically killed MOND in the eyes of most researchers.
As I said before, nothing wrong with writing papers exploring the consequences of MOND or trying to come up with alternatives, science does not work by consensus, but the bar for dropping Lorentz invariance is really high given the experimental support for it.
Secondly, even the proponents of MOND concede that you need to introduce some amount of dark matter in order to explain e.g. observations of the Bullet Cluster. So the competition isn't between new particle vs MOND, it is between MOND + new particle vs new particle.