I've seen it written somewhere that the maximum anticipated NV gain ever to be experienced in ideal situations is 3dBm over legacy 1x/EVDO to NV 1x/EVDO. Since the maximum threshold between LTE and EVDO is approximately 10dBm (between -95dBm and -105dBm RSSI), this will be far different than the maximum 3dBm difference between legacy and NV. So, therefore, we can anticipate LTE 1900 should offer less coverage than legacy EVDO. If you use Cloud RF to model the difference, in an ideal flat area, you get about a one mile difference on a 300' boomer with no downtilt.
So I think LTE 1900 coverage should be quite good in most urban/suburban locations, except in those places with poor site spacing (like Baton Rouge). But tertiary and rural markets will look like LTE swiss cheese until LTE 800 arrives toward the end of 2013.
I appreciate your offer of being a S4GRU guinea pig.
Robert