'Questions were raised at the Sept 6-7 ASAC Meeting concerning the Y+
configuration's ability to perform adequately at its highest
resolutions.  A draft document which addresses these concerns is found
at http://www.tuc.nrao.edu/~mholdawa/highres.ps.  Very simple imaging
cases were investigated (ie, two point sources a variable distance
apart, or two point sources with a moderately extended low brightness
Gaussian component), and I investigated different array
configurations' abilities to accurately determine the point source
positions and fluxes.  The Y+ configuration with 36 moved antennas
does not compete very well with the 14~km ring (or other Y+ options)
for sources separated by less than 20 mas (about 1.5 times the
synthesized beam).  When the point sources were separated by more than
20~mas, all of the configuration options did more or less similarly
well (though the ring array did get into intermittent trouble as the
point sources were further separated and they were sitting in each
other's side lobes).  Much of the difference among configurations at
the highest resolutions went away when I used the somewhat more
complex image structure with the low brightness Gaussian, or when I
added realistic phase errors (random residual phase errors, as might
exist after fast switching phase calibration, and a much smaller
systematic phase term).  Nonetheless, some of the criticism of the Y+
configuration with regards to the highest resolution was justified.
The 42 move Y+ configuration, with higher resolution, was more
competitive with the 14~km ring, and I will adopt a 42 move Y+
configuration.'