Sorry, just read your other post,

The test was done with the cylinder set so both the exhaust and the inlet valves were closed. So as such, any escape would either have to be due to a poorly seated valve (unlikely as car runs reasonably well), or running past the piston.

My understanding from speaking to the chap involved was that the 5% was so insignificant, that it was not likely to yield a telling result.

I will ask him if the test can be carried out with the valves open to see if that can yield a result with the inlet/ exhaust suitably blocked. I know nothing of leak down testing, so have no clue as to whether this is realistic!

JJ