Elevation is in feet (or k-feet), distance is in yards (or k-yards). Try this table I made:
http://www.mediafire.com/?pc3ufdb7dgqjdThe .xls is for microsoft excel 2007 and newer, .xlsx is for microsoft excel 2003. At a base of 0 elevation, I measured the flight time and amount of drop for the shell to explode at the every #.0 and #.5 fuse length (then ran a regression equation to best fill the lengths in between). For each of the 4 engine heavy bombers, I calculated the speed (in k-yards/second) at 5 thousand foot intervals (starting at sea level) using the AH speed charts.
Assuming the target is flying straight at you at the speed indicated by the AH chart, I determined the fuse length necessary to hit the target when it is at the fuse lengths distance.
(e.g., A set of b-17's is traveling directly at your flak 88 position at an altitude of 15k feet. When the direct line of sight distance between you and the target is 8.9 k-yards, you should fire a shell length of 7.2. Given the rate of the target coming to you and the rate of the shell going to it, the target will be at a distance of 7.2 k-yards when the shell explodes.)
The amount of lead to give the target is completely up to the gunner. It will take the shell 13.1 seconds and 1.5 cm of drop to get to 7.2 k-yards. So you'll want to aim where you think the target will be 13.1 seconds into the future, then add an additional 1.5cm.
Granted, this test was done at 0.0k at a rough barrel angle of 45 degrees. I haven't had very good luck with this system because by the time you make the setting corrections manually, the target has changed distance and your solution is obsolete by the time you fire. Not knowing the precise elevation, speed or trajectory of the target or even the angle or orientation of the gun greatly hampers your effectiveness.
I've seen some guys come impressively close to my formations, though. Even at 20k I've been able to here the puffs, but with only 1 gun firing and the large amount of error caused by the unknown variables makes this unconcerning.