I see this all the time in industrial pneumatic designs where an understanding of how cylinders are properly powered with pressure and slowed with restrictors/check valves.

The applied pressure to the in side of a cylinder should not be restricted, and it is the exhausting side of the cylinder that is where you restrict it to control the speed.
The problem comes in that the out restricted side will be needed as an in requiring pressure and must not go thru the restrictor and now the other end becomes the out or exhausting air needing restricted

In industry a cylinder flow control restrictor has a built in check valve to bypass the restriction when going opposite direction. What you described is exactly why not doing it proper does not work,

Here is a diagram of the cylinder restrictor/check valve disregard the valve porting prior to cylinder

https://fluidpowerjournal.com/determ...ylinder-speed/

I use clippard mini restrictors with built in checks for smooth retracts. Some times you do get away with restrictors in the exhausting side for up retract and none on the down side because air pressure going into a restrictor for down generally needs little air volume as the weight of the wheel and gravity work in your favor snapping down is generally better than up.