August 11, 2015: I took a little time today to see if I could eliminate vias in the design just to be sure I had a path to do so. With a few minor adjustments, I was able to eliminate all the vias. The biggest assumption is that the reprogramming connector can be eliminated. In the worst case, it could be replaced by pogo pins arranged conveniently around the board. Best case, it's not needed at all and micros can either be gang programmed before population or they can be entirely programmed through the RFID.

I have rigid prototype tags that currently measure a potentiometer, but their range is much worse than the eval kit that has a different antenna. In reading the spec sheet, I think the resonant capacitor on my prototype tags is wrong. I copied the EVM value, which was 39 pF. The datasheet tells me my cap should be about 51.8 pF - 35 pF, or about 16.8 pF. I'm going to try to change this tomorrow to see if my range improves. I only need an inch of Z height, but I also want to make sure it's reasonably tolerant of misalignment.