Wish there was a whiteboard here.
Scenario 1, shooting a target from a moving platform in a vacuum.
Bullet leaves the muzzle with an X and Y velocity, hits the target in T seconds, missing the point of aim by TxY
Scenario 2, same test in an atmosphere. Bullet has an X and Y velocity, and is slowed equally in both directions, as it's movement is the vector sum of the velocities. Hits the same point (albeit lower) than Scenario 1, i.e. TxY
Scenario 3, shooting from a stationary platform at a moving target in a vacuum. Bullet only has X velocity, target Y right angles to it. Bullet misses centre by YxT again, same as 1.
Scenario 4, same as 3, but with an atmosphere. Bullet has X velocity, target Y. Bullet is dragged, only in the X plane, leading the time of flight to be T+t. Misses centre by Y*(T+t).
t is the "delay", "lag" whatever terminolgy you need.
Scenario 5, shooting from a moving platform at a moving target. Is a superposition of scenario 2 and 4, so the bullet misses the centre by Y*(T+t)-Y*T, or Yt - the miss is related to the "delay", and the speed of the moving platform and targets, no the absolute velocity.
Scenario 6, stationary target and platform, and move the air between them at the same velocity as the elements moved in 5. Result is the same as 5, with the miss being related to the delay, not the absolute velocity.
In the case of a supersonic .22, the 1200fps projectile gets there quicker, but is delayed more than the subsonic, leading to higher drift.
This holds up to .22 magnum velocities, when everything starts to catch back up.