That's what MIRVs are for.
The strategy, if it existed at all and it isn't just a post-hoc explaination, was probably designed around the time the strategic bomber was the primary, and only, method of delivery of atomic bombs.
I did not know about MIRVs, this comment thread has been educational.
This was pretty much the strategy towards the end of the war. More than 60 Japanese cities were firebombed, civilian casualties probably exceeded those killed in Hiroshima and Nagasaki.
Even if there was an effect, I’d be very surprised if the signal wasn’t wiped out first by railroads and then by cars.
The goal was to get Japan to surrender, and Emperor was considering it. His generals were fanatically opposed, and if the Emperor were to be killed by the bomb, they would fight to the last man, woman and child.
US administration could then pick it's poison - leave Japan as is, suffer and inflict massive casualties from a ground invasion, or commit nuclear holocaust.
Playing around on Nukemap, 5-6 800kt warheads are enough to kill the majority of people in Houston and give third-degree burns to the rest. A Russian RS-28 Sarmat ICBM should be able to carry 10-15.
Or, with a single 20 megaton warhead, it would kill 1 300 000 people immediately and injure 2 000 000 people.
I do think it is an effective strategy. Look at Hiroshima and Nagasaki, tokyo and kyoto were spared for different reasons but they were spared. No sense in wiping out every single city. For the US, a nuclear attack would probably be against symbolic populated cities like new york, san francisco or chicago, ideally the US would surrender after that point and if there is no surrender, military targets would make more sense (hawaii,san diego,san antonio,anapolis,etc...).
The idea is, if you have a small number of targets to defend you can defend them better. Just like with infosec, you want to reduce attack surface so you minimize exposed services or inputs.