[youtube][/youtube]
After this less than succesful (in terms of power) experimentation, I decided to take a step back and examine what wasn't right about the design, and came up with the following:
-too little flow, stemming from a relatively small port and a stiff spring that didn't really let the piston pop that much
-a lot of air was being wasted flowing past the piston instead of through the port
I set about trying to improve the design in such a manner that the above problems would be eliminated. The first issue was simply solved by scaling up the design by doubling the port size and tripling the piston size. I also made the piston out of a syringe, that not only greatly facilitated construction but gave me an airtight seal ensuring air could only escape through the port.
As an added benefit, this allowed me to eliminate mechanical springs and simply use a pneumatic one by pressurising the area behind the piston, meaning I can easily vary strength by changing the pressure.
It moves a lot more air and there certainly is potential, note how it rocks with the pipe unsupported:
[youtube][/youtube]
I'd say the idea certainly has more potential now, watch this space
Here's what it looks like and a rough diagram of the internals:
I'll pay more attention next time 
