The system I'm currently leaning toward consists of 16 Sharp 230W panels and a Sunny Boy 3kW inverter. Apparently, due to what's called stringing limitations, I cannot install a Sunny Boy 4kW inverter with 16 panels. If I chose 14, 20 or 22 panels, it would work. That changes my expansion plans.
One suggestion would be to install 16 Enphase micro-inverters instead of a central inverter. The microinverters would allow each panel to turn on individually, instead of having to wait for the array voltage to reach a certain level before a central inverter turns on. Microinverters could possibly boost output 5-10% because the panels will generate sooner and stay on later and are supposed to work better in shade and on hazy/cloudy days. Also, if one dies, the rest of the array will still produce. However, there appears to be a $2000 premium based on my system size and they are relatively new technology (but have a 15 year warranty). There is also the issue of replacement in case of failure - someone would have to hike up to the roof and possibly remove panels to get access to a failed microinverter (depending on how they do the installation). I have also been told that around 10-12 years, normal inverters will probably need replacement, so I assume around 15 years, the microinverters will need to be replaced - and there is extra labor associated with their replacement. Finally, with so many microinverters, there is an increased number of points of failure vs. a single regular inverter.
Panel brands is another topic of interest. One installer is very high on Evergreen Solar and how their panels are rated 0-10% above spec vs. other brands which can be +/- 5% or so. Sunpowers are high end, high efficiency, but somewhat restrictive with their installation. Perhaps this is why none of the 5 quotes I received used Sunpower. Sanyos are also high efficiency and high priced, but are manufactured/assembled all over the place. Obviously, each dealer will tout the panels whose manufacturer they have deals with or which provides them the largest profit margin.
For the consumer, it ultimately comes down to cost, wattage and size (room for future expansion) - at least it does for me. When you start to use the more exotic panels, it's hard to justify the longer breakeven period. I chose the 16 230W panels instead of 20 175W panels due to density - I want to have room on the roof for future expansion. The 20 175s would nearly occupy the entire back roof. I could get 16 Sanyo 215W panels for a $2000 premium (and 240W less output) vs the Sharp 230W. The Sanyo's are about 8" narrower so I could squeeze 2 extra columns - or 4 more panels - in the space of the 230W Sharps. Overall, I could squeeze up to 27 Sanyo's vs 20 Sharps - 5.8kW vs. 4.6kW output - a pretty large difference.
I am leaning toward the 3kW inverter at this point. In the future, if we need more capacity, I could then go with microinverters for the additional 4-6 panels that there's room for. With today's panels, I have room to go from 3.7kW to 4.6kW, but in 5 years, panel density will be better - so I should be able to get more power out of the remaining roof space.