Only the water in the mix 'boils' out of a leak? 'dunno;
Are the Dex molecules somehow larger than water, and only the smaller water molecules leak past an open spot?
Now, as the coolant mix passes thru the hottest part of it's journey (piston cylinders), the water part can separate as steam, and is mixed back as it's cooled down in the cycle. This is why SCAs are added in to protect against cavitation in wet sleeved Diesels, giving the nitrite protection against the mini explosions occurring with the steam formation.
As steam, perhaps only it leaks past an open spot?
During the class action suit brought many years ago, with only some small gasser engines of the late 80s/early 90s, GM did some extensive testing to find the truth behind what was happening internally to these engines.
What they found was, the owners never serviced their cooling systems and they ran 'em chronically at low levels, with air displacing the lost volume. Over time, this void left many areas unprotected against metallurgic reactions of dissimilar materials.
They summarized that the most important thing was to keep the system topped off, and the second being diminished quality of the coolant. (see pH test below).
Knowing this^, I can see how you might draw the conclusion that internal damage stemmed from too concentrated of a mix. When in reality, it was the loss of overall volume and poor quality that did 'em in.
Add to this that along with switching to green, you also have a shorter coolant lifespan and is changed more frequently, adding to the Dex myth that it's a terrible thing to run and that green is better.
Had you serviced the Dex as frequently, you wouldn't have seen any problems either.
Keep the pH in check, and keep 'em full, is the key with Dex....or for that matter, any coolant mix.
What is the proper method using multimeter?
Pay no attention to the opening print statement, it's wrong. Whatever level on your meter your choose, affects where the decimal point lies. But this shows the gist of it.