Thanks both of you. After trying to come up with some sort of cat's paw clamp to reach in there and clamp the guides, I finally tried pretty much what Brian suggested. I cut the end of the Van's pins off, tapped for the 5/16" - 24 setscrews, and put a drop of paint in the end of each setscrew, which then marked the location in the door frame quite nicely. After drilling 7/16" at the mark, I test fit the door without the guides and only had to make slight adjustments; it was easy to not the direction of the adjustments. Then I used a pin for alignment and, pushing in the direction of my adjustment, clamped the guide in place for drilling. My door surface is perfectly flush at the latch points. I did have to grind down the outboard surface of the guides by about 1/16" (less, actually) to allow the door to seat flush. I suppose I could have avoided that if I had set the guides on the door itself inboard just a hair instead of flat on the inner surface.
Installing the angled pins was very easy but I had to tap them about 1/8" deeper because the setscrews protruded from the pin tube too much even though I tapped to the full depth of my tap. The finished product is super tight and very effective. I had to be very careful with length because I'm using the flush latches so I only get 90 degrees and to get full engagement of the pin in the closed position means I am very close to dragging the pins in the guides in the open position. I won't know for sure until I get the flush latch mounted but if it is a problem I may notch the outboard edge of the guide very slightly to allow the pins to pass. That would still leave the meat of the guide cone to pull the door in and of course the engagement area in the guide and fuselage bulkhead would be unaffected. I did some pull tests on the door and no slipstream at any reasonable speed will affect it; the Plane Around parts are amazing.
Speaking of that, I did try to latch the doors with only one pin in place (with no pins in, it is very apparent that it is unlatched). If the latch is closed to lock, the single pin engages so well that I cannot pull the door out or in the direction of the disengaged pin to the point where the door will open. At most, I could open it about half an inch, which I would expect to create a severe knocking in flight. Moreover, it took effort to get only one pin to engage; I had to deliberately set it at an angle to get it to happen (I could see that happening if a seat belt or strap were inadvertently outside). I'm not knocking the building or operational abilities of those who have had problems with the doors but I find the latch much more secure than the finger latch in my -6A tipup. I think the discussion regarding full engagement of the pins through the metal bulkhead members is right on. That said, I will remain cautious when this project finally moves into the flight phase.
Oh, and regarding the strut attach bracket, I didn't know it matched the hole pattern exactly but I was able to match drill #22 without damaging the threads and then enlarged the holes to #12 with no problem. My question was more regarding the clamping surface area of a bolt versus a countersunk screw in that area. So far, one person says screws are fine, so I may have raised a non-issue.