[evlatests] Closure

[Rick Perley]

    I forgot to note in my previous message that the 'closure' errors in
the L-band data were negligible -- below 0.5% and 0.5 degrees -- for
all antennas (EVLA and VLA).