20140409: It has been almost two months to this point. Probably six or more of those ten weeks were spent working on my day job, and another two weeks waiting for the boards to arrive. All this leading up to the moment of truth...
First I wanted to have a control test to prove that my development system could talk to a known good board. I ran my old DbgOut Test Probe project on a trusty old MSP430 Probe board and proved that I could program and single step the board. OK.
I grabbed a simple blink project from the TI Code Composer samples, making sure it was configured for the MSP430F1121.
There was a brief period of despondancy when the LED remained dark and I could detect no signal on the oscilloscope. Then I realized I had made a stupid count-by-zero error in my 5 line program. (Never too simple for a bug!)
SUCCESS!! I am able to program and single step the EZ430 board! The diodes do not light up, but I suspect they were placed backwards.
SUCCESS!! The diodes were backwards. I flipped the pick direction 180 degrees, built another board, and the LEDs are now a'blinkin'!
The really cool thing is that to debug the problem I built three boards, the MSP430 on every board was operational! The turn time was about 5 minutes to apply solder, 5 minutes to build the board (seriously!), 20 to bake, 10 minutes to mount the header and jumper wires, and 2 minutes to test. All three variations were built and tested in a span of less than three hours. Without access to my own build machine, it would have been an exercise in frustration to build the boards by hand (I can't even see the LEDs without a microscope!) or paying hundreds of dollars and waiting two weeks for each build from an assembly house.