The approach I've used for grouping them (even if I can't make an exact measurement)
uses an SWR bridge and a dummy load.
Wind a sample coil on the core, put it in parallel with the dummy load, and measure
the SWR on one or more bands. Generally, the higher the impedance for the same
number of turns, the lower the SWR will be, so the higher the permeability.
To adapt this to various core sizes you can pull up the Amidon datasheets here:http://www.amidoncorp.com/specs/
and compare the permeability of a specific type (perhaps 43 for ferrite and 6 for
powdered iron) to see how it changes with the core dimensions. So if, say, your
standard is 5 turns on a half inch core, you can then determine the equivalent
number of turns to try on other core sizes.
You can calculate the expected SWR for a given inductor in parallel with a resistor
for a specific frequency, and from that estimate the inductance.
It isn't a perfect approach, but by measuring at both 80m and 10m (and perhaps
VHF) you should be able to roughly group the cores by permeability.