I am pretty sure that actually the pilot when aiming saw the circle in the centre. Imagine if he really had to lean sideways to to see the full circle. This basically is equivalent to have an aim that is offset by the amount by which the revi is offset from the centre. Consequence: Your convergence point would be hence offset too in the lateral direction in order to meet where the sight indicates it would meet.

Even if this offset may be small it would be awkward and I find it hard to believe that one would like his convergence point out of the symmetrical plane.

(the situation is a bit different for the colimateur sights of first world war fighters where they had to lean to the side at some point. This however never prevented the convergence set to the centre and one's aim was just by aiming through one gun the other following accordingly. This is thanks to the way a classical colimateur sight works. This is not the way a reflector sight works.)

What made Oleg and his team to do it this way and not another? Probably because of the same reason why in old IL2 in particular and occasional in CoD some things are not there or do not work as they should: By lack of information of how this works. Just see how much discussion is going on on this subject. And some stuff may seem odd at the first glance and be dismissed immediately just because of this.

I understand that introducing binocular view in a monocular world is a challenge but I think a solution can be found. I believe in the creativity of the dev team.

Yes, swiss, the current view is in wide field like looking with both eyes at the centre of the windscreen while leaning back. The gunsight view is like leaning to the right side and forward in order to stare with both eyes through the gunsight. I have difficulties to imagine that pilots did that in a dogfight.