Actually Cam it does not tell you what your margin of error might be. GDOP, geometric dilution of precision relates solely to the angular differences of the various LOPs. That is what your unit is showing you. It is not showing you such things as the effect of transient noise on each individual LOP. Anotherwords, the computer is doing the thinking and evaluating for you. Similar factors are what make your shallow water alarm go off consistently in 500 fathoms of water and yet not utter a peep as you gring your way ashore.(g) Similar technology is what makes your automatic acquisition radar track minute squalls but manages to overlook a sailboat 5 miles away. Again, I am not trying to be critical of GPS, but rather the manufacturing process that leads to, "they don't need that" mind-sets.
Perhaps an illustration would make the point better. "Transit" was generally employed with a gyrocompass input and a manually entered speed input, what with very few ships having pitot logs. Theoretically, if the speed entered was off a significant amount the receiver should not be able to acheive a "fix". Nevertheless, the receiver would merrily pump out positions, some ten miles off, as if all was well in a carefree world. Advances have certainly been made, and the receiver's estimate of GDOP would reveal the potential for large errors, if monitored. But the unit still does not give the operator the type of information he needs to either accept or reject individual LOPs. "You are here" logic is the quite predictable result.
“Scientists are people who build the Brooklyn Bridge and then buy it.”
Wm. F. Buckley, Jr.