The idea that the two parcels of air must meet again at the trailing edge has no logical basis. It has been dismissed long ago by most people who know anything at all about aerodynamics.
It might be better to consider what the air must do in order to permit the wing to move through it.
Let's imagine a wing that is 30 cm thick. As it moves forward, the air in contact with the upper surface of the leading edge must move upwards and the air in contact with the lower surface of the leading edge must move downwards to make a space 20 cm thick for the wing to pass through.
If the air is to remain in contact with the wing, then after the thickest section of the wing has passed, the air must move back to close up the gap.
The acceleration needed to achieve this depends upon the amount of time taken for the wing to pass through. This time is determined by the TAS and the chord length.
For any given TAS and wing thickness, increasing the chord length will reduce the acceleration rate that is required. This will reduce the highest local mach number for any given free stream mach number, so MCrit will increase.
So to have a high MCrit we can use very thin wings, and/or very long chord lengths (such as with delta wings) or sweepback the wings.