Thanks guys for your helpful tips and good ideas. I understand your feelings that it MUST be a leak. But it is not. Yes, this is counter-intuitive: It runs with the throttle completely closed and blocked off. The intake has been off three times and proper gaskets and proper sealant used. The list of tests [everywhere there is a joint] for leakage is long and comprehensive (including all of the suggested methods from you guys, plus many you have not). There is no leak. Many knowledgeable racers/engine builders have been to the shop to help find the leak. You can imagine the hair pulling I, and they, have gone through - the bafflement - the doubt - the impossible to accept notion that it is running with no leak and the throttle blocked shut. . . . but it does. Yes, I know what you are thinking now: the tests were not done right, the reassembly was incompetent, the materials used were faulty, something was not checked. I have not been casual in my approach to finding the solution, nor cavalier in my efforts; many, many hours of painstaking effort has gone into checking for leaks in a very systematic manner - without finding any. I am still open to any other ideas.
The alternative reasons are radical indeed: too much overlap or bad valve seals/worn valve guides (yes, I know, there should be white smoke from the oil if it is the valves . . . .). I am checking the cams first (yes, it is possible, and documented, that an engine can run because of misindexed cams), and then a leakdown test (I have to borrow a leakdown test kit first). I'll know in a couple of days. Wish me luck.