The SGHWR, possibly the least-pronounceable reactor design of all time, was actually fairly similar in concept to the Advanced CANDU Reactor currently being promoted. I don't see the attraction myself, the combination of heavy and light water seems overcomplicated, but then I'm not a nuclear engineer.
Some of these designs are really just difficult to wrap your heads around without a diagram of what is going on. I found one for CANDU that shows the light and heavy water circuits.
In this diagram, the heavy water/moderator circuit is yellow. The light water/coolant circuit is blue and red.
I'm not aware of there being any advantage to using heavy water as the coolant instead of light water, so using light water in the coolant circuit seems like an expedient way to save money.
Deployment of the SGHWR was planned for several sites in the late 1970s, but by the time new reactors were being ordered new British designs were out of favour and AGRs were ordered. Subsequent reactors would have been PWRs - the CEGB liked the reliability and standardisation the American designs could provide. The UKAEA made a big deal about PWRs being unsafe and (especially early on) unproven by comparison to 'their' gas-cooled reactors.
I talked to my professor some more about this (he's a British nuclear policy expert), and there are some interesting insights regarding this.
In the legislation for the second cycle of the British nuclear power program (1964 to 1974) it was required for liquid cooled reactors to be developed and deployed. The Advanced Gas Cooled Reactor actually is liquid cooled, because it uses liquid carbon dioxide as a coolant.
Gas cooling also has some inherent safety relative to liquid cooling under certain circumstances. In a gas cooled reactor, you can heat the gas all you want and it remains stable. In a water cooled design, you must maintain a safe operating temperature or the water will convert into steam, which has reduced cooling capabilities as it is a gas. If the temperature gets too high you run the risk of it undergoing spontaneous electrolysis, producing helium and oxygen. Apparently there isn't much of an explosive risk from that (most explosions are due to steam, the same as any boiler), but it does create a fire risk that can lead to other issues. Three Mile Island and Fukushima suffered from spontaneous electrolysis.
Gas cooled reactors have the issue of Wigner radiation from the graphite moderator they must use. Wigner radiation is energy that gets trapped in the graphite, and it can result in power surges when the control rods are inserted into the reactor. Windscale suffered from Wigner radiation.
Now, the worst type of reactor you can have for safety is a water cooled graphite moderated reactor, as with the RBMK of Chernobyl infamy. That combines the worst attributes of both types, because you can get spontaneous electrolysis and have to worry about Wigner radiation induced instability. When Chernobyl suffered cooling problems, the water began turning into steam, leading to a temperature increase. This led to the fatal mistakes that happened next, and arguably best practice would have been to do nothing. Instead the operators put the control rods back in to the reactor, but that caused a power surge and panicked the operators even more. They decided to try to put more water into the reactor to cool it, but that just ended up turning into steam and eventually undergoing spontaneous electrolysis, leading to the fires and explosions.
The reason why Windscale was able to be recovered partially through the use of water is because it was a gas cooled design, so it was under low pressure to begin with. There was still a risk of spontaneous electrolysis though, which is why that was only resorted to when the fire began to approach the temperature rating for the concrete and the structure was at risk of failing. I'm not sure if the water actually stopped the fire or if finally shutting off the air intake did, but the water injection and closing the air intakes stopped the incident.
One final consideration for gas vs. water cooling is that gas, or at least carbon dioxide, tends to corrode metal. This is an issue since irradiation already causes strange effects such as crystallization.
Fort St. Vrain in the United States was helium cooled and had corrosion issues as well, but that was due to water leaking into the helium circulators.
I'm rather curious as to what the British Phase III HTGRs based on the DRAGON reactor would have been. I've found hints that a demonstration plant was planned for Bradwell, or maybe Oldbury, and probably would have been a 625/660 MW(e) plant, but very little besides. I think this might have used closed-cycle gas turbines, but that's little more than speculation.
No argument from me that British nuclear engineering would have done a lot better without the five (!) consortia trying to outdo one another in the engineering stakes.
Perhaps it would have been similar to
Fort St. Vrain. With a Brayton cycle closed cycle gas turbine it could have even higher efficiency.