Yep, and this was at a time when JCS was expecting it to be not noticeably worse than Okinawa (40,000 KIA, 150,000 WIA) for BOTH Olympic and Coronet and estimates gave the Olympic landing/follow-on force a numerical advantage.
I never understood where that estimate came from or the methodology behind it. In fact, if we actually go through the math it suggests that there would be 2,000 more KIAs but 20,000 fewer wounded on the Kanto plain than in southern Kyushu. It looks like someone's wild guess written on the back of a napkin in order to placate Truman at a time when the Army was concerned the President's support for the invasion might be wavering.
On the other hand, if we look at the Army Service Force's actual planning documents from around that time it's immediately obvious they expected things to go much differently. From "
History of Planning Division, ASF" vol. 7, part 7 - "Redeployment of the United States Army after the Defeat of Germany" (15 January 1945):
Replacements for Battle Casualties - The estimated average rate of replacements required for dead and evacuated wounded during the 18 months of the redeployment period is 43,000 per month.
- p. 330
This is for the Army only during the period from July 1945 to December 1946, and presumably doesn't include wounded who could have been treated in-theater and returned to their units within the 30-day (later 120-day) limit.
In part 8, from pages 372 to 374 there is a detailed breakdown of projected Army strength in the various active theaters vs. Japan, and page 391 gives the replacement and rotation schedule from the mainland United States. If I've interpreted correctly, after taking rotation into account what's left over are replacements to cover 267,000 dead and missing and about 598,000 WIAs evacuated to the US through the first part of 1947. Of these, 224,000 dead and 499,000 evacuated wounded would have been prior to the end of 1946. This matches well with D.M. Giangreco's quote that in the Spring of 1945 ASF General Somervell was working with figures of "approximately 720,000 replacements" for "dead and evacuated wounded" through 31 December 1946.
Furthermore, the ASF estimates - in addition to not saying anything about the other services for obvious reasons - appear to operate on the best-case scenario that total forces for Coronet would not have exceeded 25 divisions: neither the 3 division theater reserve is shown leaving the Philippines nor the 17 division strategic reserve in the US. Though, the 11 December 1944 version of Coronet (
Volume 6, part 1, p 28) made it clear that up to the full 45 division force (2 million men) would be committed, comprising 2 Army Groups and 5 Armies.
As for the bombs, while their historical use was a moral atrocity, I cannot see any American politician of the time period deciding not to drop them.