No. This doesn't sound like a method that reads through the skin. What it sounds like to me is a system that uses a plastic 'strip' that has a reservoir for blood. The blood (probably a big drop) goes onto the strip, and the strip goes into the machine. Somehow, through some kind of refractive sensor (an optical sensor of some type), the clot, or some other marker, is detected. This sounds somewhat similar to CoaguChek, which looks at impedance changes (IIRC) rather than formation of an actual clot (which Coag-Sense does).Which requires blood to leave the body, be exposed to a trigger and clotting time measured. Do you understand how that is measured?
You seem to be talking about a magic method of determining this under the skin.
I'm still unclear about how a clot can form quickly enough for the user not to just get tired of waiting for something to happen (if no reagent is used), or how INR is calculated if there's no divisor. Perhaps they'll figure out some implicit standard that is characteristic for their method of testing that can be used to calculate INR.
I'm not going to theorize about how they'll actually make the leap from clotting time detection and INR, but it's clear that there WILL be a drop of blood that is used for testing.