MOND is just some wild idea, but a little thought should convince every physicist of its uselessness. It has major issues both in explaining experimental data and in its theoretical consistency. It justifiably receives next to no attention from the vast majority of (astro)physicists.
In popular science the idea however does not seem to want to die, perhaps because it is so easily explained to a layperson. Of course this is a little frustrating for the community, but perhaps we should look at the upsides: more attention for science is probably a good thing, and explaining to people why MOND is so useless provides a good opportunity to discuss some proper physics.
Sometimes you gotta be wrong before you get it right.
I mean, Newtonian mechanics are "wrong" but served us well at some scales for a while, and that it observationally failed in others led us to relativity. Even "relativity" took iterative steps, from Poincaré's Lorentz invariant theory (or even earlier with Galilean relativity) all the way to GR via special/restricted relativity, the latter name having been retconned because it's only valid in restricted special cases and fails to unify generally. And we know GR fails to unify with quantum mechanics, so one of them (or both) gotta give.
So even if something as MOND were "wrong" and known to be wrong (definitely so), there's still value in experimenting with it to get a better understanding of things. That's just how things work.
I disagree: some experiments are just not worth our time. I wrote about such a situation three years ago:
https://news.ycombinator.com/item?id=26656206
My view is that it applies here as well.