Although the water heater is pressurized the same as the water in the pipe, I don't think of public utility water as "high pressure", in the sense that the water would have to reach the critical temperature before there was enough pressure in the pipe to force water from the hot water heater back into the supply. Water boils at around 320F under 60psi pressure, and 60psi is more than I get from my utility. Once the water had boiled locally, the copper pipe would have been free to warm up rapidly as there would have been nothing left to absorb the heat. Since the copper is able to stand 60psi it wouldn't have ruptured and there would have been no reason for the water to flash since it would have been below the critical temperature (around 700F).
Another thing is that the mass of water in a 1/2" copper line from the W/H isn't all that much. Compared to the mass of water in the tank itself, heating the water in the pipe would be easy given enough time.
One test of this hypothesis is whether or not the fire started at a localized high point in the pipe, such as where it came out from the top of the W/H, went into the wall, then went back down into the floor. If vapor was forming it would have collected in the highest point local to the source of the vapor. If that's also a location that was being heated, eventually the vapor would have filled the pipe, the liquid would no longer have been there to absorb the heat, and the temperature would have spiked.