I built a solar hot water system for my father's house. To save money, I built my own solar panels using copper tubing, wood frame, cover glass, and styrofoam board insulation- two 4'x4' and one 3'x3' panels. Built a south facing platform sloped for the latitude. Modified the hot water heater for solar, and surrounded it with additional insulation. Installed plumbing, check valve, differential controller and recirculation pump. The system was not very efficient, but it worked.
Later I designed a system using double-walled transparent polycarbonate glazing panels with manifolds glued onto the ends, water blackened with india ink, a heat exchanger and 250 pounds of paraffin wax for heat storage in a 55 gallon oil drum surrounded by fiberglass insulation. The sun heated the blacked water directly; the pump circulated this water through the heat exchanger counterflow to the domestic water. This heated the stagnant domestic water in the heat exchanger to 137F buffered by the melting point of the wax. When a hot water tap was opened, this preheated water flowed into a conventional hot water heater. As the wax solidified it supplied its latent heat of fusion to the flowing domestic water. Design flaws compromised the efficiency of the system; however, it worked, and I believe a commercial unit could be designed to be entirely satisfactory.
If I were to do a domestic solar hot water system today, I'd do it differently yet again. I'd use a high efficiency commercial panel- 4'x8' or 5'x10'; Large pre-heat tank of perhaps 160 gallons; small hot water heater of perhaps 30 gallons; hot-water/cold-water mixing valve. The solar panel would preheat the water to whatever temperature the conditions would allow. If this temperature was lower than required, the water heater would raise it. If the temperature of the water in the heater was higher than required then the mixing valve would add cold water to the outflowing hot water to reduce the temperature to that specified. This would entail no penalty because it would merely cut down the flow of water from the heater- thus conserving hot water in the pre-heat and heater tanks. I'm certain a system could be properly engineered for hot water requirements and climate to provide a high percentage of total yearly domestic hot water needs at reasonable capital expense.
A fourth alternative is to have flash heaters and mixers at all hot water taps, and no conventional hot water heater- only the solar pre-heat tank. This would have the highest capital costs, but also would be most efficient in operation.
Interestingly, Israel has the largest usage, proportionate to popuation, of simple thermosyphon solar domestic hot water systems in the world