You can't just suddenly change the meaning of a blue check from "this person is who he says he is" to "this person paid $8" and not expect it to blow up in everybody's face. Maybe there was a non-terrible way to implement that idea, but this should not have been done.
I'm sure there's a lot of learning happening, and Elon is clearly not afraid to blow up expensive rockets in pursuit of learning, and that has worked out pretty well. But there is no need to light buildings full of money on fire to learn easily predictable things, or things which almost anyone could have told you if you'd asked.
Most people on Twitter by now reflexively see a blue check as meaning the account is verified to be under the control of the brand represented in the account's name. That's the entire controversy here. The blue check used to have that meaning, and now it suddenly does not, making it potentially useful to buy a blue check for the sole purpose of impersonating a brand and using the account to harm that brand.
Does the new bluecheck offer preferential algorithmic treatment? That was never clear to me. But, even if it does, is it worth the potential for (say) me to create an account called @McdonaldsOfficial, pay $8 for what recently implied I really am McDonald's, and tweet something enormously harmful to that brand? Again keeping in mind what just happened to Eli Lilly, which was exactly this.