Hi Chris, again, the advise you've got is all good, but in truth, you need do nothing about the length, you can leave .015 dead space at each end, which I would recomend over leaving it all on one end, or you can simply leave the piston .030 thicker/longer and allow it to take up the extra space. The hardest way would be to increase the stroke when you machine the crank, so it has an extra .015 in stroke, and as John said, make notes on all the different things it would effect. You should have enough rod length to allow for adjusting the piston to equal travel and equal head space on both dead centers, or if necessary, add the .015 to the rod length, it should be threading into the cross head anyway, as you always have to adjust the length of the piston rod to set up your dead centers (top dead center and bottom dead center). The rod from the valve eccentric also is adjustable with threads and a lock nut I assume, and the rod running through the valve should have a nut on either end of the valve to set the timing, so if you have to make the valve rod fifteen thous. longer, no big deal. The main issue is you've got the port centered, and everything else is just individuality in your engine, nothing more or less. You can even claim you did it on purpose to get more horsepower, and draw diagrams demonstrating the truth in your statement.
As to machining the valve face, I would bolt the angle plate directly to the table, as you are "stacking tolerances", in clamping it in a vise. As a rule, you will tend to be more accurate, the more you work directly on the table of the mill, and the fewer parts of "jig" between the part being machined, and the table, assuming the mill is fairly accurate with respect to flatness of the table, perpendicularity of the spindle, and more importantly, perpendicularity of your quill travel. Much error has crept into work because the spindle was indeed perpendicular to the table, but the quill travel its self was not. You can test this by putting an indicator in the spindle with your best square on the table, and running the quill down and up, with the indicator against the square. If you are not certain of your square, you can do the same thing twice, once on each side of the quill, and if the reading is zero on both, your quill is indeed perpendicular, and if the reading is say .005 on one side, and .005 going the same direction with the quill i.e. down on both or up both sides, it is your square out of truth. All that said, you will have dead space top and bottom just to keep from hitting when things get loosened up a bit from running, I'd just go ahead and mill the valve face, and use the X axis for reasons stated by Nick I think.
Personally, I prefer castings to billet, as they save lots of metal and machine time when they're right, and in my experience, castings I've worked over the past ten years are substantially better than those I worked on thirty years ago, lots of them come from very high tech controlled atmosphere furnaces, being done with automated equipment, and with better quality original metal and a better knowledge of the ore going into the original foundry. You will note, most people who machine out of billet end up doing much work altering the final profile to more closely resemble castings, and how they make the whole of an engine look. As a final note, when you mill a flat at each end of the cylinder to drill your ports to meet the cast in ports on the valve face, it is worth the time and effort to move the table all around, with a drill bit in, or some drill rod, and ensure the shallow angle you will be drilling, will meet the port as clean and well as you possibly can as you will always deal with "wander" when drilling long skinny holes, but that is perhaps the hardest part of the cylinder operations, and not too hard after all. It's nice to see good workmanship such as yours, and a pleasure to watch it come about. cheers,

jack