I did a test recently -- today! -- myself on this, and found that the butterfly keyboard does slow me down a bit compared to my usual rate, which I wouldn't have actually predicted; I'm slightly slower on it in terms of raw keystrokes and also slightly less accurate. But it's not an incredible difference, and there are other keyboards I've used (like the much-hyped Brydge keyboard for the iPad Pro) which had more deleterious effects.

Interesting. I wish there could be a corpus of data on this to help understand what makes the keyboard better for some and worse for others.

I feel that I am more likely to brush a neighboring key and that key is more likely to trigger on the butterfly keyboard of my 2017 work machine than on my 2014 home machine. I end up doing a lot more corrections on the butterfly keyboard.

It seems like the tighter spacing and lower travel result in more unintentional key presses.

Yes, I think the keyboard needs the typist to adapt to it more than is appropriate for a mass market keyboard. I had the 2014 also and had to unlearn how I rested my hands on it and touched it. Not everyone will equally succeed at that, as evidenced by my coworkers who are still hitting the 2017-2019 keys way too hard.