I don't think you would get water hot enough, It would mostly like need to be high pressure steam. You might get one to run with hot water, But not enough to produce any power. They have little engines that will run off a hot cup of coffee. Some can run off the heat of your hand. But all of the low temperature ones I have seen are doing good just to run and never produce power. Thats why I really wanted to verify the 5 HP. 
Another related thought I have kicked around: Use wood gas and an O2 generator for a hotter more concentrated flame to the engine. Not heating up all the nitrogen in the air should add efficiency also.
I saw a lot of them on youtube also, I always thought they were just a scientific curiosity. But check out this quote from wiki (from the link above)
"In the conversion of heat into mechanical work, the Stirling engine has the potential to achieve the highest efficiency of any real heat engine. It can perform theoretically up to the full Carnot efficiency, though in practice this is limited. The practical limitations are the non-ideal properties of the working gas, and the engine material properties such as friction, thermal conductivity, tensile strength, creep, rupture strength, and melting point. The Stirling engine can run on any heat source, including chemical, solar, geothermal........."The efficiency of the typical listeroid is what, 20 to 30 per cent? I know I saw a figure somewhere but can't remember exactly what it was. This was the only thing I could find on wiki on real world efficiency numbers for the stirling
"Compared to the idealized cycle, the efficiency of a real engine is reduced by irreversibilities, friction, and the loss of short-circuit conducted heat, so that the overall efficiency is often only about half of the ideal (Carnot) efficiency"which would come to about 30%, correct?
And this
"To summarize, the Stirling engine uses the temperature difference between its hot end and cold end to establish a cycle of a fixed mass of gas, heated and expanded, and cooled and compressed, thus converting thermal energy into mechanical energy. The greater the temperature difference between the hot and cold sources, the greater the thermal efficiency. The maximum theoretical efficiency is equivalent to the Carnot cycle, however the efficiency of real engines is only a fraction of this value, even in highly optimized engines"In other words it runs on the temperature difference between the two ends. The reason I threw the 1000F/200F question in is because if you had 1000F going in (easily achievable with sunlight) and all your waste heat at 200F, the coolant temperature would be hot enough to still be useful. Granted, you'd lose some efficency but if you'd send all the energy through the stirling first, and then use the rest in a CHP setup, a couple % difference in efficiency wouldn't matter all that much.
I thought about the O
2 generator already too, around here WMO would be used instead of wood and temps on that would go up to 1000F+ I would think....especially if you used one of these burners and used O
2 instead of air
check out the babington burner, very simple design, clean burning, might be some possibilities there........
http://www.green-trust.org/2000/biofuel/babington/default.htmhttp://www.aipengineering.com/babington/Babington_Oil_Burner_HOWTO.html
