In order to propel a pellet to appreciable velocity, you tend to need around 100psi sustained for most of the barrel length (assuming a reasonable length around 375mm).
When you combust a butane-air mixture in a closed vessel starting at atmospheric pressure, you can acheive a pressure around 100psi, but that pressure will decay as the pellet travels down the barrel . As the pellet takes it's trip down the barrel, the gas space volume between the pellet and the combustion chamber increases. Some leakage around the pellet also contributes to some pressure decay.
With a very large combustion chamber like the Tippmann PEP, you've got a lot of initial volume starting at 100psi so the pressure doesn't decay very much as the ball flies down the barrel.
However, I was hoping to fit my combustion chamber in the slim profile of an 870 type shotty. Because of this, I couldn't fit the chamber that I needed. Furthermore, said chamber would have to be elongated and on the narrow side. This is not conducive to very rapid and complete combustion. You get the fastest flame propagation in short squat chambers (closer to spherical shapes is better if the spark plug is at the centre).
Conversely I was working with long tubular chambers of comparatively small radius wrt to length.
I tried to beat this problem with a high pressure setup. I was basing my designs around a pump action shotty so I figured that a bit of elbow grease could precompress the air-butane mix to say 2-3atm (30-45psig). Precompression would greatly increase the peak pressure (linearly related to atmospheric pressure ~15psi) and I hoped that it could reduce some of the chamber shape problems. Higher compression ratios crunch molecules together and greatly speed flame propagation.
The problem which finally stymied me was designing a seal which could be airtight yet survive combustion temperatures. Butane flame fronts can reach 1900C which is well beyond what organic elastomers can durably take.
I think the PEP operates with the mix initially at atmospheric pressure. Because of this, you wouldn't have problems with the fuel mix leaking past the ball very much. You could use the ball as the forward plug to prevent gas from oozing out the barrel.
At one point, I thought about mixing in a bit of sacrificial liquid like water into the butane reservoir. The idea was to spurt a little liquid onto the oring which would evaporate and absorb heat so the oring didn't actually get singed. You can observe a similar effect by soaking a piece of paper in ethanol. Light it on fire and the paper doesn't burn until all of the ethanol evaporates. Evaporating parrafin prevents the wick on a candle from burning as fast as a cotton string with no wax would burn.
My test combustion chamber could be prepressurized, but the movable seal on it wasn't durable. Liquid oring preservation didn't show to be reliable. It wasn't possible to get a consistent wetting of the oring. Silicone oil would have left a more lasting film, but I was concerned with the combustion products of silicone oil. I didn't really want to invent the first airsoft gun finally deserving the designation of "Cancer gun".
The reason I was looking at butane instead of propane, was that a smaller volume of butane was necessary to hit the ideal mix fraction with air. It allowed slightly more air (20% oxygen) in the combustion chamber and consequently more energy. More carbon dioxide produced = more energy req' more oxygen.
With the exception of gas mortars and odd funny guns, I couldn't shoehorn a combustion system into an airsoft gun. I did work out a pretty cool ignition arrangement for a shell loaded nerf mortar though.
Unfortunately liability reasons have convinced me to shy away from productizing a mortar launcher and product development of a mortar within a 750sq ft apartment is rather annoying.
Want nearly free GBB gas?