The plane does NOT have to be on or even near the “arcs”. The plane can be anywhere within the area circumscribed by the arcs, which represent the horizon of the satellite, which was located above the Indian Ocean at the location marked on the illustration in Post No. 4.
I initially thought the same - but other reporting (and the search area) have indicated that the satellite ‘pings’ could determine range. Somewhere on here I even saw a graphic which showed the two potential ping locations at each of 6 pings.
I will admit that it looks suspiciously like each ping gave a nearly identical range to the satellite, which seems very unlikely...I assume (I hope) that a ping from the 1st hour of flight was used to validate the accuracy of the ping range finding. It seems the searchers have placed great faith in this, as they are only searching the Indian Ocean in the vicinity of the last ping.
I’ve also thought they could do a degree of elimination, by determining which satellites did NOT ping the plane...assuming there is overlap in coverage, the circular coverage zone of the one satellite that did ping it could be reduced by the overlap zones that did not ping the plane.