From: Nova Scotia originally, now Michigan
I've been curious about this for a long time so spent a few hours today running some tests. Using the Air Campaign introductory scenario I sorted and organized my Air Groups using the Turn 1 advice from HarryBanana's Guide to Strategic Bombing. I kept the ADs simple, 1 strategic Bombing of Essen, area 3, for Bomber Command. 2 Strategic Recon ADs for 8th Air Force, 1 to match the BC bombing raid and 1 of Hamburg. Then a strategic bombing of Hamburg for the 8th. Finally a Strategic Recon of Vienna and matching bomber raid for the 15th.
I ran 6 tests, 3 times each, varying only the Strike Num and the Req AC in the directive advanced options.
I set the number of strikes to use all of the bombers available/divided by the Req AC. So, for example, 1,400 bombers available divided by 100 per strike equals 14 strikes.
Here are the average Bombing VPs for each size of strike:
Test 1 used the AI (Auto)defaults: 375
Test 2 used 100 bombers: 508
Test 3 used 200 bombers: 476
Test 4 used 300 bombers: 440
Test 5 used 50 bombers: 571
Test 6 used 25 bombers: 604
Test 6 generated 600+ points every time I ran it.
It certainly looks from this like smaller strikes generally produce better results, certainly substantially better than the AI defaults, which all the help posts point out, but also appears on the surface to be much better than 100/200/300.