I like your noise and vibration dampening material - it's actually quite effective 
Added some rubber feet and with lots of fettling it's quieter and much less inclined to run off!
There were a few problems which were causing the engine to randomly stop, I think the main one was the alignment of the valve guides/stops with the valve chest and the way I chose to fix them. I enlarged the holes so they no longer act as guides to keep the rod on course but merely as stops. The length of the valve itself does a good enough job of keeping the rod straight. Fixing them with a single screw through the base isn't a great design - if they work a little loose with vibration they can rotate and cause the rod to rub. That said, now that there's no contact between them and the rod the loosening seems to have gone.
I also thinned the crank by about 0.25mm as I think it was pushing the con rod outwards slightly. I decided to do away with the top bars of the cross-head as they were serving no useful purpose and creating more friction. You might see in the video that the gland nut is jiggling about as it has no packing yet.
I also think it could have done with a bigger flywheel: if I try to run it with the lowest pressure I can, the crank alternates in the same semicircle, there's not enough momentum to get it over the hump. Even so, it runs the same in each direction on very little pressure - in this video the compressor output is reading zero throughout and I turned it down until it stalled at the end.
I think that concludes this build (other than maybe applying a bit of Loctite to stop things working loose and filing down the screws at each corner).
