It has been my experience that EI meters can be fooled by surface moisture, concrete density, steel reinforcing, and concrete constituents among other things.
While some of these factors may not vary widely within a single pour of concrete, think how they can vary between a slab poured in Northern Michigan compared to one in Southern Florida.
EI meters measure how well a signal flows through concrete. Dense concrete will transfer the signal more readily that porous concrete, steel contributes also. Moisture is what we are looking for, and moisture helps the signal along, discerning how much of the reading is moisture and how much is due to other factors is the key.
My beef is the inspector who walks onto a job site, plunks down his EI meter and says "Five! It reads five so the slab is too wet, sorry."
Well five what? Percent? I just took a reading of my fireplace hearth which is limestone and it reads 0%. I don't believe it has 0% moisture. Then for fun I placed the meter on a piece of steel and it read 7%. While the steel I'm sure has no moisture in it, the conductivity of the steel is very good and read the equivalent conductivity of concrete at 7% moisture. I did this just to demonstrate how materials can effect the readings.
Now move to a job site. How do we compare a 12" slab to a 3" slab? Wire mesh reinforced or maybe not. Maybe even steel fibers. I was on a job where the 5/8" rebar was sticking out of the surface of the slab in dozens of locations. What will that do to an EI meter if we place it near rebar that is just under the surface cream? I can tell you the slab will read wetter than it should. The surface of a slab may be more porous than the concrete an inch or two down. Or the opposite can be true. How are we to know?
So do I think EI meters are worthless? Not at all. I can quickly walk a floor and take EI readings of the surface in literally hundreds of spots. The concrete density probably doesn't vary as much as it will from multiple batches poured under varying conditions. The constituents should all be the same. The reinforcement will vary, but I can work around that as I am reading. What I'm looking for are areas with overall higher or lower readings to conduct follow up testing to determine cause. It will not tell me of moisture problems deep in the slab unless they are bad enough to show at the surface. Then I need to think if surface moisture or temperature may be throwing my readings off.
You can imaging how concrete at dew point reads very high. You may not see surface moisture because of a slight breeze from HVAC or natural convection. The concrete can darken from moisture but still have a dry surface. Readings again will be artificially high.
There's a lot of thinking involved when using EI meters, I just touched on some that came to mind. Sorry for the long rambling...