I knew I was gonna open a can of worms. The filling process cannot be explained by saying it is two sides of the same coin though... that somehow drawing a column of air is the same as it being pushed. There is no drawing of air.
The piston does very little work while the motor is filling. Its swept movement increases the volume of the crankcase, which is almost immediately being filled at near the same rate as its volume is increasing, by pressurized external air, in accordance with its port timing events, port shape and port volume. Filling is initiated as a response to piston motion. Very little stretching of air molecules is taking place inside the crankcase, so the piston does not have to do much work during filling. A true pump must perform all of the work required for filling. So there is nothing to be gained by thinking of a motor as a pump.
Nitpicking I know... but look around and you'll find many, many folks who maintain a false pumping model in their minds. While modding for performance those folks typically resort to brute force tactics over time, attempting to create different variations of a pump... which can lead a person to repeat past mistakes while becoming one-trick ponies... because they cannot divorce themselves from false pump model concepts.
Can 'o worms, those damn motors are, regardless. But if you're gonna brute force a strato motor to behave as a conventionally scavenged motor, you're probably better off to begin with (or convert to) a conventional design in the first place. Got lemons...? Make lemonade.
Hope I'm not coming across as being a dick head. I'm not very good at this stuff.