Let's test some 6DJ8's in order to sort out the meter resistance.  

Here's the data collected in prior log.  This is a dual triode 

Pinout:

The letters correspond to how the tester refers to the pins.

The tube data lists these values:

Type 2 means, a B+ of 30v, plate R of 820 ohm and meter R of 360.

Test circuit

In order to  test, I'll take the fact that the meter+R2 much higher than the 40ohm of the pot, and omit it initially.

With this, I'll use the 200ohm pot connected from power supply negative to cathode, and 820 ohm set on 10W decade resistance boxes from plate(plus grid) to the power supply positive.  Then we'll be able to measure a voltage across the power supply negative and pot wiper.

With a 200ohm pot value of 21/100, this means 42ohm from low to wiper.

Measuring that voltage represents the voltage across R2 and the meter resistance.  Knowing that a good value results in 0.5mA, we can calculate the total resistance.

Each line represents a different tube, measuring that voltage, along with the current through the circuit.  This is just from the simplified circuit for the emission test. (Not open, leakage or short tests)

I'd prepared to test more, but stopped after 8 because things were so clear.  There's even a failure of element 2 in number 4.

One surprise is that the current is so high.  The 6DJ8 is specified at 27mA max, with 15 typ.

That the voltage comes out to just over 1V doesn't seem like a coincidence.

If we take 1v as passing (above 0.5mA, this means the resistance needs to be less than R = V/I = 1/0.0005 = 2000.  Removing the selected series R we get a meter resistance of 1640 ohm.  

If the meter were actually 1.5K, then the resulting current would be 1/1860 = 0.54mA, which seems like it may be in the good territory on the display.

If the meter were actually 1K, then the resulting current would be 0.74mA, right in the middle of the good side.  (I notice that in the instructions for testing new tube types, it indicates a be default reading right in the middle of good, so 0.74mA makes sense)

Looking at the triode with an off value, with a 1K meter, it would be 0.64mA, probably reading good.  With a 1.5K meter, it would be 0.47mA, probably in the bad territory.

Seems like we've narrowed it down in that 1 or 1.5K region.  Given that, we can confirm that omitting it from the circuit was reasonable, reflecting only a 2 ohm difference.  At 1% of full scale, i can tell you on this pot, that this is well within tolerance of manual setting of the pot.

This all seems to suggest to me that about 1.0V is that passing target.

With a little uncertainty about the threshold and test current, it think I need to test another model of tube, with different settings.