Interesting thoughts, Nathan - and although I have to admit a fair bias toward 'navigation' error that puts Gardner (Niku) in the relative crosshairs (likely by accident, not design) I remain fascinated by the various navigational possibilities.
Logic and puzzle solving are essentials, of course. Best also I think to look at the simplest and most likely scenarios - the more linkage that gets piled on, the less likely success in proving the null, etc.
Maybe that is why I tend to gaze at dry land (coral) more than open sea: on one hand, the open sea is dreadfully simple - it is HUGE, and the gut screams that they simply never found land and piled in, out of gas; on the other, crews of land planes tend to navigate with land in mind and wish not to ditch, and the size of the ocean drives the mind toward alternate land possibilities.
Where are the odds on all that? In my mind, such as it is, 'land' is a high draw; open sea is to be shunned. For argument's sake though, consider the size of the sea and assume for a moment that's where they went: your point is apt - where then to look? One has to mindfully 'channel' Noonan's thinking to sort that one out, and I personally believe (subject to being wrong, of course...) that a large set of variables lie there - it's not a matter of flying out to the third white cap and turning left 90 degrees and holding for 12 minutes.
More seriously, I can appreciate the search patterns that have been postulated, but less so than I can the employment of the LOP and eventual outcome of land found in the Phoenix group. We at least do have the tangible 'on the line' last call, as cryptic as it is.
And we may just have ever so much more. Let the chase continue.