It just niggles me a little that if you join those "last ping" arcs the curve goes through the last active reported position S of Vietnam to a reasonable accuracy. Yes I realise if the aircraft had sat ditched and powered for seven hours something should have been found. So not a very plausible scenario. But still it niggles.
Me too. It does niggle.
Presumably the SAR operation would not have been activated until the aircraft was overdue Beijing at 0630, plus some slight margin. I believe that's the norm. Sunrise wasn't until shortly after 0600 that day, anyway. The operation had been activated by the time of the MAS press release at 0724, but the last "ping" was at 0811, little more than three-quarters of an hour later. They'd have been very, very lucky to have overflown the incident site in the first hour or so of the search, even if they were airborne and on station by then. Yes, I know the sea area is shallow, but human beings miss things, and stranger things have happened.
Publicly at least, the abandonment of the search east of the peninsular seems to have been driven by the Malaysian military radar contacts, which the authorities weren't convinced by for several days, hence the divided search.
I hope they were right in their eventual choice.