It's interesting, because the old "linear" mode seems technically correct, e.g. a clip that's -6dB looks like it's half the height (half the volume) of 0dB one. And the one -12dB is half the height of -6dB, and so on. Whereas the new "perceptual" mode the clips get progressively less tall, but that simple relation isn't there anymore. I'd even go as far to say that they should name them the other way around, as the old/linear shows change in waveform height the way we perceive the loudness...
LINEAR:
PERCEPTUAL:
But I'm not complaining - I thus far avoided working especially with bounced audio, because I'm typically somewhere around -24 to -18dB on a track, so I couldn't see shit there.
Great job by devs, especially because we could avoid the confusing data zoom from Studio One: