That "potentially leaking exhaust valve I feel might/could degrade throughout the test and could change the result in the end ruining all the results (I just have two words in mind when thinking of compressed flamable gasses: Cutting Torch)
#1 (x) HP,
#2 (x) HP -3,
#3 (x) HP -5,
etc ??
It may not be a bad idea to run a compression test following each first manifold swap to see if the compression is holding or is dropping thus maring the results. The worst I ever saw was a large portable industrial compressor with a Rotary Screw Compressor and a Detroit Diesel for power. The unit developed an internal fire and majorly cut through the 3/4"-1" Cast iron compressor housing oil mist 130PSI and burned the compressor to the ground. Further I am sort of surprised that you are not repairing that compression mess prior to running things (the manifold switcheroo you are doing is seriouis work and removing a head or two, dropping them off at a shop when you are that far into the motor is opening risky.
I am not sure of the speed that valve burn occurs but I've seen engine blocks ruined by poor sealing at the cylinder head and a (cutting torch) burn through two adjacent walls. I have no idea how long this burning takes and at least this could doccument the time of this action.
I agree if the compression holds the test will be great I am hoping things do not burn.
