There's a data set, in the base.
There's a data set, in the rover.
The base to rover processing via rtk, is the same data set used via ppk. The METHOD is not the same. But, the data set IS THE SAME.
Thus, it's possible/likely that if rtk is bad, then so is ppk.
Now, change the data set, and we're on to a genuine check. Without that, you **REALLY** don't have independent verification.
It's only maybe a 10% (or so) independent verification. Because ppk handles the data somewhat differently.
I guess that's kind of what I was getting at - in most cases, they're either both right or both wrong. So having an RTK and PPK coordinate for the same point isn't really a "check" of anything, since both coordinates are of equal quality/reliability. (assuming similar epochs/seconds for RTK and PPK)
If you leave your LS for a long observation (20 min plus) and you get 3 or more epochs with 300 plus separation, then you have the right init.
Then, if ppk is within a tenth, it'd be a pretty solid observation. Use ppk for a little higher accuracy, and rtk for verification.
Very rarely would this be wrong.
Being pedantic, I'd do it 2x though!