On casual inspection, I accepted the forums dogma that the canard on the J-20 is plane-aligned to the opposite wing, and that the ventral fin on the J-20 is plane-aligned to the opposite tailfin. However, doing measurements, it's obvious that this isn't actually the case.
On the preceding image, we have wing measurements:
The line formed by the wing is 36x348 pixels, giving us a ratio of 0.10344827586206896551724137931034 or about 5.9377724516054286242797916868402 degrees.
On the subsequent image, however, we have canard measurements:
The line formed by the canard is 61x332 pixels, giving us a ratio of 0.18373493975903614457831325301205 or about 10.587385565866989310103810801075 degrees.
That's a difference of roughly 5 degrees, and too great to be the result of measurement error. A measurement of the opposite canard from a different picture, shows a greater difference of 6 degrees, but the canard in that picture is not in that resting position. It is possible that the canard in the original picture isn't in rest position either, but the difference is only 1 degree based on minor disalignment.
This is something reasonable to assume based on the J-20's engine deficit; i.e, the airframe can't fully align the canards with the main wing without causing significant penalty to the canard-wing aerodynamic alignment, or further reducing the effective wing area on the J-20 due to increasing the anhedral even further.
The next set of measurements would be for the ventral fin vs the tailfin alignment.
Here, we have 179x345 pixels for the tailfin. This is a ratio of 0.51884057971014492753623188405797 or 31.25 degrees (measurement error would allow +-1 degree).
For the ventral fin, however, we have 104x211 pixels. That's a ratio of 0.49289099526066350710900473933649 or 29.5 degrees, or in other words, there's a 1.75 degree difference between the ventral fin and the tailfin.
So what we have to conclude is that the J-20 does enjoy some plane alignment features, especially when viewed from the z-axis, but on the y-axis the set formed by the canard-wing planes and the set formed by the ventral-tailfin planes aren't internally aligned.
This isn't a big problem insofar as with WS-15, it'll be possible to rework the aircraft, especially if the ventral fins can be entirely replaced by TVC.