There is much to say about vacuum
tube quality and testing but without going deep into theory &
technical details, here are my comment on this issue:
One should always check for shorts and internal element leakage before evaluating the quality of a tube. If a short or serious leakage is present, discard the tube and DO NOT PROCEED FURTHER (serious damage to the tube tester may result)! When it comes to evaluating a tube's quality. The way to do so is to measure the Gm on a tube tester that designed to measure such. An additional test (significant for some tubes and in some circuits) is a test for gas. Diodes and voltage regulators can't be subjected to a Gm test.
New tubes (just like used ones) will vary considerably in their Mutual conductance (measured in micromhos) or Gm (for short). Test results will vary depending on the type and brand of tube checker. Emission type checkers will only measure emission and this reading is very unreliable and will vary significantly with various tube tester models. When testing rectifiers and diodes there's no such thing as a Gm test. Diodes/rectifiers make poor amplifiers.
So with this in mind, when a seller states a tube's quality in terms of; "like new", measures "NOS" or "near NOS" (or in terms other than "Gm") is not a knowledgeable seller. I'm not saying they are dishonest, but I'm stating that there is a lack of understanding of their product or the proper way to evaluate same.
Standards were set up long ago as to what the minimum acceptable tube test readings should be (and the readings will vary with brands and models). Some tube testers will list the average, others; the minimum acceptable or expected readings. These readings are explained and specified in the testing data for the particular model tube tester in their manual.
I use a Hickok model 539C, considered buy many to be one of the most accurate and desirable testers available (sans laboratory models) the posted values on it's charts are; the minimum acceptable.
Here's an example:
The minimum acceptable reading for a #27 tube is 630 Gm. If you have one that tests 1100 Gm, that's well above the acceptable reading and is about as good as this particular tube will measure (NOS or otherwise). A tube testing a bit below the minimum (or expected value) is not necessarily a candidate for the trash can, NO! it may work just fine in some circuits, so don't throw it away.
Keep in mind almost none of the popular tube testers that were sold and used in the service industry are not 100% reliable and (including the Hickok 539C) are not laboratory instruments, all have their shortcomings. The most reliable tube tester today is a modern instrument, the Amplitrex.