The longer the focal length of the lens, the more the subject will move in the viewfinder due to the same amount of camera movement. This is the case because the total angle of view visible in the viewfinder is narrower.
If you are using a lens with a 24° horizontal AoV, such as an 85mm lens on a FF system, then a 1° angular movement of the lens' optical axis is 4.16% of the total horizontal AoV. If you are using a lens with a 10° horizontal AoV, such as a 200mm lens on a FF system, then that same 1° angular movement is 10% of the total horizontal AoV.
Movements that are too small for you to notice when using a 24-85mm lens may be just large enough to be perceptible with your 70-200mm lens at longer focal lengths.
The reason it occurs immediately after the shot is because the VR and/or IBIS systems are compensating for the movement until the image is taken. It's then probably "recentering" the VR and IBIS systems immediately after the image is taken to allow for compensation in whatever direction is needed for future movements as you take more shots.
Back in the old pre-IS days it was not uncommon at all for a shooter who had never shot at longer than, say, 50mm - because that is what most "kit" lenses were back then - to get their first 80-200mm or similar telephoto lens and complain about how blurry it was, especially at the longer end of the lens' focal length range. In the vast majority of cases it turned out be a revelation of the need for better camera stabilization technique. What small movements one can get away with with shorter lenses will be more and more obvious as the focal length increases. This was often further compounded by the fact that consumer level telephoto zoom lenses tended to be several stops slower (typically f/5.6 at the long end) than the 50mm lens (typically f/1.8 or f/1.4) they'd been using. The same amount of light required four stops longer exposure at f/5.6 than at f/1.4! Shooting at 1/250 at f/1.4 with the 50mm "kit lens" meant using 1/15 at f/5.6 with the 200mm telephoto lens in the same light! Not only were the same amount of angular movements creating more blur, but the longer exposure meant that the motion continued for sixteen times longer during the shot! Four times the percentage of the height of the image compounded by a sixteen times longer exposure meant that the same rate of angular movement would give sixty-four times the length of blur in a resulting print!
