I may as well throw in my $0.02. I am not one of the experts on this site but I have restored and sold around 25 002A and 003A generators, with NO complaints. I kept two 002A's and two 003A's for my personal use. I think I know them fairly well.
A few months ago I acquired one 802A and one 803A. The 802A had been "reset" (completely rebuilt to as-new condition). It looks and runs as new and has 12 hours on the meter. It still has a tag attached to the red emergency shut-off knob reminding me to change the "break-in" oil after 30 hours use. The 803A is clean but well used and came with 1494 hours on the meter. It starts and runs well but the % power meter does not work. I tested the % power meter and the meter is OK. Just to be sure I changed out the original % power meter with another one. Still no indication even though I am drawing 40 amps at 240V. Instead of doing major trouble shooting to find out why the % power meter does not work, decided to install one of the digital volt / amp meters offered by a regular on this site. More on this below.
When comparing the generators IMO the 002A and 003A are a simpler design and they are easier to work on because they are modular. The 802A and 803A are not modular and (IMO) are a much more complicated design and much harder to get to the "innards" for maintenance and repair. That is why I gave up on trouble shooting the cause of the % power meter not working and decided to install one of the replacement digital volt / amp gauges. I have installed identical digital gauges on the 002A and 003A. Doing so on the 802A and / or 803A is many times more complicated (at least the first time around) because of all the work involved to get to the "innards" of the generator (accessing, unbolting, and removing L3 to install the current transformer and reinstalling same is a bear). The 802A an 803A are MUCH quieter than the 002A and 003A. One of my 003A's has the ASK which makes it much quieter than the "open" 003A without ASK, but no where as quiet as the 802A or the 803A. Since I live in the middle of 700 acres and have no neighbors, noise is not an issue, but it may be for some urban dwellers. All four generators will produce rated power until the cows come home. However the 002A and 003A definitely have a lot more reserve capacity than the 802A and 803A.
From my measurements either of my 002A's will load test 8000 watts resistive for 30 minutes with no problem. When I increased the load to 8500 watts the engine noticeably labored but generator did not quit but after about 5 minutes the engine overheat sensor shut the generator engine down. This was in 90F weather. Since the 002A is air cooled it probably would have load tested 8500 watts OK in winter weather. One of my 002A generators will start and run a 2.5 ton heat pump while the main breaker on the other one will open before the heat pump starts - go figure. Either of my 003A generators will load test 13,500 watts resistive for as long as you care the run them or until the water boils. I use water heater elements immersed in a plastic 55 gallon drum of water as my load tester. Both 003A's will load test at 15,000 watts for a couple of minutes without any complaints. I think the 003A's would load test a bit higher than 15,000 watts for a short time, but that is all the water heater elements I have.
My 802A will load test 5000 watts all day. It seems happy load tested up to about 6500 watts, which is about 130% rated capacity, but any more and the engine seems to labor and it starts emitting noticeable black smoke from the exhaust. I did not try to increase the load beyond 6500 watts. My 803A will load test 10,000 all day but it starts to noticeably labor at 12,500 watts, which is about 125% capacity.
From my tests all four generators will provide rated capacity for weeks or months at a time, but the 002A and 003A seem to have a greater reserve capacity for overload and are much easier to access and work on. Of course I do not recommend overloading any generator but that is the results of my tests.
I hope this helps.
