I usually rescale the maps by using inter/extrapolation, however you need to know the hardware very well to be able to tell if the new values are sane / optimal or not. Ultimately the only way to tell is logging.
I use the 1.8T engine with K03-05x (S) turbocharger as an example.
The maximum flow of the turbocharger is around 165g/s.
The original LDRXN map is set to limit the air charge to 114% (~0,44bar) @ 6520rpm.
Since the original rev limiter is set 6800rpm and the last entry in LDRXN map is at 6520rpm, the value used for 6521-6800rpm is the value pointed by the last know sampling point (i.e highest rpm = 114% / ~0,44bar).
At 6520rpm with 0,44bar boost the engine consumes air around 138g/s (1139g/M³ @ 37C = 7,24M³/min).
Since the same limiter is used at 6800rpm you need to know the air consumption @ 6800rpm with 0,44bar to see if the hardware is up to its job: around 143g/s (1139g/M³ @ 37C = 7,55M³/min) of air will be consumed at 6800rpm / 0,44bar.
Since the maximum flow of the turbocharger is around 165g/s and the engine only requires 143g/s @ 6800rpm we can tell the stock LDRXN limits are sane!
You just need to make sure the maximum air charge at the last sampling point (6520rpm or what ever you set it to) can be achieved at the rev limiter too. However if you only change the last sampling point from 6520rpm to e.g 7000rpm the gap between the last and the previous is likely to become too high (you need to calculate if it does).
If the last two sampling points were 6000 & 6520rpm and you only raise the last one to 7000rpm; the air charge indicated by 6000rpm sampling point will be used between 6000rpm and 6999rpm. The original gap was 519rpm now it is 999rpm. Is it still possible or sane? This is why you need to know the hardware you are playing with.
LDRXN is the easiest map to rescale frankly as there is very little data that needs to be worked on.
If you raise the rev limiter you most likely want to rescale the ignition angle maps atleast.
This is because at the higher rpm you need / want more ignition advance.