There still seems to be confusion over the 40° elevation angle arcs accredited to the last 'ping' timed by the Inmarsat IOR at 64°E over the equator.
The graphic below should provide insight into how that elevation relates to any angle subtended from the earths center in any plane. Note the graphic doesn't indicate the top or bottom of the earth sphere is North or South, as it could equally be 154°E or 26°W. Also, from 35,768km above the equator, the full disc is not seen, as reference to the graphic will show. 90° elevation is on the beam centerline, but at an earth elevation of 50° the angle subtended from the center of the earth is 35° and the angle at the satellite is 5° of its overall half beam-width of 8.7°.
When plotted on a transverse Mercator chart/map, the ring drawn will start being circular immediately below the satellite, but as the angle gets bigger the Mercator projection makes the projection elongated in a polar direction. If drawn on a equidistant projection, the circles would remain just that.
Those who have referred to the AMSA search graphics will note that the tracks supplied by the NTSB are based on two different ground speeds, and they reproduce the diverging tracks. The track indicating the faster speed is the westward one.