Estimates for the casualties of the Soviet invasion were similarly catastrophic, and I don't know how exactly you reconcile the reluctance to surrender with both bombs being dropped and a Soviet invasion incoming with the idea that the early stages of a Soviet invasion would have been enough to force capitulation. On the other hand, if your argument is, unlike Hasegawa's, that a full-bore Soviet invasion is what would have forced a surrender, you are likely correct but are looking at casualties comparable if not exceeding the severity of the planned Operation Downfall.
The delusion that air power alone could force a surrender was very common in the air force during and after WW2, but as Vietnam showed, it's not actually a viable way to force a country to surrender. The 1946 analysis, though wildly optimistic, still presumed a heightened firebombing campaign which, may I remind you, had already claimed more lives than both atomic bombs combined. And if you believe that napalm doesn't melt childrens' bone marrow and cause thousands to die slow deaths of chronic illnesses and injuries, I have a bridge in Brooklyn to sell you. The nuclear fears in our society lingering from the Cold War have more of a hand in that view than the actual facts of what bombing entails - nuclear weapons are uniquely terrible for their scale, not the cruelty of their effects. Terror bombing is cruel enough to match or exceed nuclear weapons; it just takes more resources from the attacker. Furthermore, the USSBS reached conclusions about the success of strategic bombing campaigns that are highly questionable, and very much connected to the fact that post-war budget cuts were on the horizon.
If you think it's a warcrime, if you think Truman is a monster for it, that's fine. But you have to accept what the alternative is, and own up to that. If you are determined to think that the use of even one atomic bomb was wrong, inherently, then you are faced with the unavoidable verdict that doing the right thing does not always save more lives in the end.