The ID of a 22mm pipe tends to be closer to 20mm.
I guess it depends a bit on manufacturer, but the ID tends to be very slightly elliptical - however, my personal reckoning* is that the official 20.2mm (22mm, minus two 0.9mm walls) is a good average.
*I filled a 10cm section of pipe with water, weighed it to a hundredth of a gramme and then found the average cross sectional area from the known volume and length.
If a pipe cutter was used, that might be the source of a deceptive measurement - they crush the ends of the pipe in a bit.
I almost always use a hacksaw over a pipe cutter these days, as it means I don't have to ream the end of the pipe out again. Far better anywhere flow or being able to insert full diameter projectiles/pistons is needed.
As far as your piston having a 1mm gap? Hell no. As tight a fit as possible where it'll still easily move.
What concerns me is that a barrel support like that design will probably rip off on the first shot, when the pressure in the valve falls.
Because the flow is dependent on pressure differential, the flow will only be equal between chamber to valve as from valve to barrel if the valve is at a considerably lower pressure than the chamber. Hence, there'll be no small force on that support. It needs to have a far higher flow area so the flow between chamber -> valve flow will match valve -> barrel flow at a far lower pressure differential.
The equal area rule is really quite a fallacy. Thing is, flow is defined by pressure differential.
But when flow is not negligible, any restriction causes a pressure drop - hence reducing the available pressure differential at the next restriction and thus the flow as compared to if there wasn't a flow restriction upstream of it.
(And, inversely, having a flow restriction downstream bounces back and reduces the pressure differential at the previous restriction).
The electrical analogy would be like adding resistors in series. Every resistance adds up.
The small resistances (like the natural friction of straight pipe) or even medium resistances (elbow fittings and the like) aren't all that significant when there's a big one (the valve), but other big resistances will have a major effect on the current.
I'd suggest it'd be better to have a separate barrel support to a design like boogieman/advancedspuds did
here, with lots more flow area - and instead increase the sealing face edge separately with a considerably less drastic increase in diameter.
, 
