Years ago, an energy engineer I knew did a comparison of an actual building's energy use with Carrier's HAP, Trane Trace and Market Manager.... He evaluated the time and the difference between the actual and estimated energy use. His conclusions were approximately that it took him a day to input with the Carrier HAP program and get results, somewhat less with the Trane program and a few hours with the Market Manager program. The initial results had the Trace program within 5% of the actual, HAP at 10% and Market Manager @ 20%. The interesting thing was that the Trace run and the Market Manager runs where quite similar, only the order of magnitude difference. The HAP run overestimated the cooling energy in the summer and understated electrical consumption in the winter.
The conclusion, find a tool and become intimate with it. Do lots of runs, the best tool is worthless in unskilled hands, and even the worst tool can provide valuable insight if the user is practiced in using it.
By the way, that particular engineer now uses DOE2, not because he thinks it models significantly better but because it is recognized by the federal and some state governments as well as the local utility.
I have been using Trace 700 as well as System Analyzer (Trace-light) for a little less than a year and I am getting good results but I learn about limitations all the time (often the same sorts of limitations that I had with DOE2). The point is getting that experience to know what you don't know. GIGO. For scoping work the System Analyzer is quick fairly accurate (+/- 5%) and tells me order of magnitude so I don't chase dead ends. I would be more concerned with getting support and finding a product that is stable and has on-going support and releases revisions on an on-going basis. (I will not go into the status of DOE2 and or Energy Blast, Market Manager et. al. as far as revisions are concerned).