There has never been an improvement in weapons comparable in degree and in sudden impact to the atomic bomb, General Groves explains (in Now It Can Be Told: The Story of the Manhattan Project):
In the case of other developments, such as explosives, the airplane, the tank, long-range artillery, armor-clad warships, submarines, and even rifles, it took years, if not decades and centuries, after their first use for their revolutionary influence upon warfare to be felt. In the case of the atomic bomb it took only a few hours.
[…]
When I told Arnold there was a chance that we might not be able to fit the bomb into the B-29, no matter how hard we tried, he asked me what I would do then. I said that if the B-29 could not be used, we would have to consider the use of a British plane, the Lancaster, which I was sure the Prime Minister would be glad to make available to us.
This brought from him the characteristic reply that I had hoped and expected to hear: that he wanted an American plane to deliver our bomb, and that the Air Force would make every effort to ensure that we had a B-29 capable of doing the job.
[…]
Fortunately, as time went on, we were able to make changes in the design of both types of bombs, so that it became possible to fit them into a specially modified B-29.
[…]
The effects of General Arnold’s wholehearted co-operation became very evident when elements of the Air Force appeared reluctant to furnish the necessary number of B-29’ s. Their reaction was quite understandable, for these planes were in such short supply that it was impossible to give the crews that were to operate them overseas even the desired minimum of training.
After exploring the situation, Wilson told me that the only way we could get the planes was for me to go see Arnold myself. Even then, in his judgment, we would not get them until shortly before the actual operation.
General Arnold did not hesitate. He fell in with my request almost without discussion, without any evidence of disinclination, and without any suggestion that I might get along with a smaller number. This was typical of all my experiences in dealing with him. He fully realized the importance of the project and never expressed the slightest doubt of our ultimate success.
As I left Arnold’s office to walk down to Wilson’s to tell him how I had made out, he met me in the hall and said that Arnold had just telephoned him. When he added that he was completely surprised and quite impressed by Arnold’s prompt agreement, I replied that I was not the least bit surprised; it was just what I had expected.
Not only did Arnold’s action in this case provide us with the planes we so urgently needed, but it indelibly impressed upon all his staff that MED requests were to be granted without argument.
This stood us in good stead until a few months before our actual operations against the Japanese began, when Parsons reported that he did not think our planes were in the best operating condition, and said they should be replaced by new ones. Investigation showed that he was right, but again Wilson felt it would be impossible to do much about it because of the tremendous demands for B-29’ s in the Pacific Theater.
When I appealed to Arnold, however, his response again was quick and emphatic. He said that in view of the vast national effort that had gone into the Manhattan Project, no slip-up on the part of the Air Force was going to be responsible for a failure. He then asked me how many new planes I needed, and I replied that as a minimum I would need one to carry the bomb. While several more to carry instruments and make observations would be desirable, they were not essential. I made it clear, however, that no matter what else might suffer, we must have one plane that would be in absolutely perfect working condition when we were ready for the final take-off. Immediately Arnold said that he would order fourteen new planes for us, and fourteen more to be placed in reserve to meet emergency needs. He repeated that no matter what else might go wrong, no one would ever be able to say that the Air Force did not do its utmost to support the Manhattan Project. In this, he was entirely correct.
The new lot of B-29’ s was delivered to Wendover Field, Utah, during the spring of 1945. These planes had fuel injection engines, electrically controlled reversible propellers, and were generally much better than their predecessors, particularly from the standpoint of ruggedness.
[…]
Tibbets had been the Operations Officer of the 97th Bombardment Group in the North African and European Theater of Operations, where he had flown the usual number of combat missions, and had then been returned to the United States. Since his return, he had been engaged in testing the B-29 and in formulating the instructions for its use in combat. He was a superb pilot of heavy planes, with years of military flying experience, and was probably as familiar with the B-29 as anyone in the service.
[…]
Insofar as possible the group’s officers should have been men who might reasonably be expected to remain in the regular service after the war. We should have recognized the importance of this but, as far as I know, nobody did. Although this mistake made no difference in the accomplishment of our immediate goal, in the postwar years it has been most unfortunate that we have not had in the regular service as many men as possible who were experienced in the use of atomic bombs in actual war. Indeed, sixteen years after Hiroshima and Nagasaki, only four—Tibbets and Ashworth, a weaponeer, and the two bombardiers, Ferebee and Beahan—remain on active duty.[…]
The total authorized strength of the 509th Group was set at 225 officers and 1,542 enlisted men. It was a completely self-contained unit, including besides its Heavy Bombardment Squadron, a Troop Carrier Squadron and all other essential supporting units.
[…]
In September, after it became clear that we would use both a gun-assembly bomb (the Little Boy) and an implosion bomb (the Fat Man), 4 we decided to freeze the external shapes of the three models then existent—one Little Boy and two Fat Men. This was to permit completion of the necessary modifications to the B-29 so that the training of the 509th would not be delayed. The first planes were ready in October and were delivered to Colonel Tibbets at the Wendover Army Air Base, which went under the code name of “Kingman” and sometimes “W-47.” The ballistic tests were begun that same month and were continued until August 8, 1945, by which time each of the two models finally adopted had been dropped in combat.
[…]
In December, the 393rd Squadron was sent to Batista Field, Cuba, for two months of special training in long over-water flights. The program also included training in high-altitude visual and radar bombing. On these practice missions, formation flights were habitually avoided and the crews grew accustomed to operating singly. This was not because we anticipated sending unescorted single planes against Japan, but rather because we were not sure that the escort planes would accompany the bomb-carrying plane all the way from its take-off to the target and back; therefore, we wanted it to be fully capable of independent navigation.
This special training proved very fortunate indeed when later General Curtis LeMay adopted the plan of using a single plane for each bombing mission. This placed all navigational responsibility on the navigator of that plane and was completely at variance with the standard Air Force navigational procedures. Normally, bomber missions were flown by massive formations, with the lead plane carrying a thoroughly competent navigator. The navigators in the other planes were not as a rule nearly so experienced. Fortunately, because of our insistence upon a high state of navigational training throughout the 393rd Squadron, the 509th Group was not caught short when LeMay’s proposal was put into effect.
After the 393rd Bombing Squadron returned to Wendover from the Caribbean, its training continued, and the fliers gained much valuable experience in the course of the ballistic testing of dummy bombs similar in dimensions and weight to the atomic bombs that were eventually used. At first the dummies were inert; later some were filled with normal high explosives. They were never, of course, loaded with any fissionable material. Most of our ballistic testing was conducted at a range in the Salton Sea area. Out of these tests came the information we needed to aim the final bombs accurately.