The J20 is a continuation and refinement of the aerodynamic concepts developed on the J10.
The main reason ventricle fins are added to a design is to improve directional stability, especially at high speed, and could allow for smaller vertical tails as a result.
I could be wrong, but I also believe that the ventricle fins on the J10 and J20 can aid in the high alpha handling of the plane.
A canard-delta generally has better nose pointing ability than a conventional layout design because with canards, you are directly pushing the nose of the plane instead of pushing the entire plane to turn the nose - the movement arm is much shorter, thus the same applied force would yield a far greater turning force, with mass the same, greater force = more velocity => faster nose turning. Basic physics.
However, if your tail start sliding out, the situation is reverse, thus canards are not as efficient in controlling the back end of a plane compared to a conventional layout design. This would be exacerbated in high alpha maneuvers, when the wing and fuselage of the fighter would block airflow to much, if not all of the vertical tails, thus removing the primary means to control directional stability on a tailless canard-delta in level flight.
This is a very preliminary assessment, and may be affected by many other factors I do not know about, but on the face of it, I would suspect that the J10 would have better sustained turning and energy conservation properties compared to the likes of the Typhoon or Rafale that does not have ventricle fins. The flip side would be that the Eurocanards should have the edge in instantaneous turning capabilities.
I would also expect the J20 to have similar good high-alpha maneuverability and controllability, which would incidentally, be in line with the very rudimentary assessment some model makes reached when they built a RC J20 model.
As has already been pointed out, the designers at CAC did not miss the opportunity to make additional use of the ventricle fins, and have cleverly worked their design so that the ventricle fins block direct line of sight to the engine nozzles from the side-on aspect of the J20, thus allowing it to have a far bigger stealthy aspect even with conventional engine nozzles.
Looking at the picture you quoted, would it make any difference what the actual shape of the engine nozzles are from that angle?
I am unfamiliar with the effect the ventricle fins might have on drag. Would it be higher because there are more fins, or lower because the reduction in size possible to the vertical tails because of the addition of the ventricle fins is larger than the additional drag created by the extra fins?
My gut tells me that the ventricle fins and smaller vertical fins would allow for lower drag, both because the smaller fins would benefit from the low drag shock cone effect (is that what it's called?) caused by the nose and body of the plane, and also be because having fins on both the top and bottom of the plane makes for a much more even distribution of drag. With only vertical tails, all the drag would be on the top of the plane. This would push the nose up, and force the FBW to pitch the plane's nose down to maintain level flight, which causes more drag.
With fins top and bottom, some of the nose-up force is cancelled out by the drag on the ventricle fins pulling the nose down. This would this require less of a pitch-down movement from the other control surfaces and thus would help to reduce drag. But, once again it's all down to the sizes of the different forces. If the combined drag from top and down fins is greater than vertical fins and pitch-down, ventricle fins are more draggy and less if the situation was reversed. Maybe the only way to know for sure is to stick different designs in a wind tunnels dn test them out. But on the face of it, I think the J20 is less draggy for the ventricle fines.
That's my understanding of it anyways. Maybe someone more familiar with drag can help me out there if I got anything wrong.
.
