The major determinant is the cars you work on, or more specifically, how much curvature the cars you work on have. The more curved the body panels, the more likely you are to see pad stalling on free rotation DAs. In that case, despite the longer stroke of the 15mm-21mm models (from any company), their ability to continue to rotate the pad decreases severely. This is where the forced rotation polishers (again, from any company) will actually be faster than the long stroke free spinning DAs. If you are, for example, polishing a 1970s slab sided Cadillac, you will probably do very well with a long stroke DA. If you are polishing out a Porshe 911, or an old VW Beetle, then the forced rotation DA will probably be faster, as it will keep the pad spinning.
It is possible to polish even the 911/VW Beetle with a free rotation DA, but you need to be so careful with the amount of pressure you are applying, and so cognizant of the pad spin, that it can make the polishing process a real ordeal, and it will take a long time to do it.
On a vehicle with flat surfaces, the long stroke DAs will polish faster, due to their larger area of coverage. The forced roation DAs will also not have a problem with the flat surfaces, it will take longer due to the shorter stroke of the current crop of forced rotation DAs (stroke is 8mm for Flex XC3401, 5.5mm for Makita PO5000C and Rupes Mille).
Note that modern cars are tending more and more towards curved surfaces, the exception being some pickup trucks and SUVs.
Well thats a nice theory mut not true at all. Again there are only very very few spots where you have problems with 15-21 mk2 machines. If you have problems, then you just cant use the machine. For those areas rotary or Rupes mini comes handy. And its not like 3401 is perfect for small areas either, so most of who does this must have two machines.
Me and my friend did white 2008 997 Porsche which had very hard clearcoat and had lots of swirls and scratches. He did the other side and i did the other side. I was using Rupes 21 mk2(both 5" and 6" bp) and Rupes mini with Meguairs and Lake Country microfiber pads. He was using Flex 3401 and a mini rotary with yellow hex logic(one of most agressive foam pads). We used the same compound which is Nanolex heavy cut(better than Menzerna 400), i would guess onbar with Jescar). I corrected my side much faster. He tried my Rupes and admitted that it actually does cut faster with much less effort. Again Porsche is very curvy car and i didnt have any stalling problems. That friend is going to buy Rupes 21 mk2

Now if someone asks that why didnt he use microfiber, well because microfiber dont really work all that well on forced rotation da. Not on Flex 3401 and not on my Rupes Mille. Even forced da with something like Lake Country purple foamed wool doesnt cut like Rupes and microfiber. The thick Buff and Shine microfiber is ok with forced rotation, but not great.
Oh and the cutting power of Rupes is not because you get larger are covered but because you get larger motion. For example Rupes 21 mk2 with 5" plate and pads corrects quicker than with 6". So yeah.
here are couple of examples of Mk2 performance and atleast on that 21 mk2 video there is no washer mod and with it the stalling is even less of a problem or non existing
https://youtu.be/N74Tj9WRf4M?t=114
https://youtu.be/cBWkYIYzhes?t=167