Looking at the residuals (GridLeak plots), I can see that things are similar to pp200, witth what appears to be an offset in GridLeak distortion at all z (our current understanding of the distortion has it go to zero at the ground planes, |z| = 208 cm). I include here three plots of GridLeak distortion residuals from tracks for high luminosity pp400 data (runs 6154024-6154031, bbce+bbcw > 2e5):

[radius on horizontal axis, distance from central membrane |z| on depth axis]

We can see here that standard pp200 corrections are insufficient, but the 30% increased corrections aren't far off.