Hmmm, that's kind of similar to what I did (I can snap a picture in a bit), though my linkage is a little different. The first attempt I had it up as far as it would go against the idle cable stop (which I know was enought to stop the engine before, but maybe it was not stopping the fuel entirely). After that try, I took the cable off and put the stop linkage all the way up, so the lever sat against the valve cover. I'm not sure if it was still delivering fuel at that point, but the first try would have been enough to lubricate the cylinder (and comes out the GP holes) on subsequent tries I guess.
I didn't think unplugging the fuel line into the IP would help because the fuel left in the IP would still enter the cylinders, wouldn't it? Even though it's not much, it would do the same thing that happened last night, and provide a little lubrication and give high readings.
I'm going to put it all back together and run it a bit to get out any fuel that may have sat in the cylinders overnight. Then I guess I will experiment. What happens if it fires on the cylinder you're testing? I know it won't start because the other cylinders can't fire, but wouldn't that be pretty damaging to at least the tester?

__________________
1987 300D, arctic white/palomino--314,000 miles
1978 240D 4-speed, Euro Delivery, light ivory/bamboo--370,000 miles
2005 Jeep Liberty CRD Limited, light khaki/slate--140,000 miles
2018 Chevy Cruze diesel, 6-speed manual, satin steel metallic/kalahari--19,000 miles
1982 Peugeot 505 diesel, 4-speed manual, blue/blue, 130,000 miles
1995 S320, black/parchment--34,000 miles (Dad's car)