Hi Stuart, the relief valve starts as a center drilled hole from the top of the head, centered front to back, so the small tip of the center drill is all that pierces the port, and does so in the taper section. The angle of the center drill makes for a seat for the ball, it being about .112, with the body of a #1 center drill being .125. I then drilled down most of the way with about a .140 drill, leaving about .100 in of the center drill body hole intact, to keep the ball in tight, but to allow it plenty of room to let built up pressure out on the return stroke, my testing initially having the ball fly about three feet in the air from the pressure, and fortunately found. I then put the head in the mill vise cylinder side up, put a .042 drill in, lined it up on the hole where the center drill tip pierced the tapered port, and then drilled through the head, above where it seals with the cylinder, but below the edge, through the back side to the relief hole, then just about .030 in into the front side, so the pin which is going in the hole will be captive once the head was bolted in place. The ball seats very well in the taper part of the center drilled hole, and holds a vacuum when the crank is turned backwards, but easily relieves pressure instead of the valve having to flap, and the pressure built up blowing out the flame, or at least causing it to flutter and interfering with its steadiness. I've seen similar ideas with a port made, and a leaf of spring steel over it, as a reed valve, and these can be in any orientation, and hidden as well, as I've heard in some engines I've watched on u-tube, but couldn't see.
I've found the initial start up with alcohol, always leaves lots of condensed water, both inside the cylinder, and on the valve surface, which makes for lots of friction, or strong sticking of the valve, and in the cylinder, makes an emulsion with such oil as is there, that makes for lots of piston friction. In washing out the cylinder with WD-40 first, it washes out what old heated oil was there, leaves some solvent in the cylinder, and disolves most of any sticky oil, eliminating the extra "stiction", while the remaining solvent vaporizes very easily, and adds to the initial power of start up, helping the engine to start off with some speed, and build up enough heat to keep water vapor from condensing, and forming the sludge.
I'm not sure if there will be as much water vapor from the butane, as there is from alki, I'm hoping not, but if there is, I will continue to pre-clean with WD, as it has been good on both of my larger cylinder engines. My Duclos "flame sucker" runs well with a bronze piston, a 360 brass cylinder, a bronze valve, and occasional take down and scribbling with a pencil for graphite lubrication. Waiting too long on this means scoring of both cylinder and piston, but rubbing down the piston with very old 600 wet or dry, and a six by six inch rolled up piece of the same, rolled tight, dropped in the cylinder and allowed to expand, and used just to take off the tops of the scoring, about two minutes overall, and then the scribbling inside the cylinder and outside the piston, means it has another twenty or thirty minutes of good running. Scribbling with the engine assembled can keep it together for weeks, but forgetting means the engine will be running great and then suddenly stop dead, and it's better dealt with before that.
I'm pretty hot on the check valve idea, I think a lot of engines on shelves would possibly run with this addition. It completely changed the feel of this newest engine, it just needs to be kept washed clean or the ball will stick.

hope that clarifies it somewhat, Jack