Last night I remembered that there is a much better source than the rough hand-drawn sketch for where the survey points were. The end product of the Bushnell survey was a map similar to, but far more detailed than, the New Zealand survey map.
I know that it is fashionable to doubt, but my impression is that the tiny printing on the map says:
Aerial photograph assembly sheet
South Pacific Ocean
Phoenix Islands
Gardner Island
Soundings in fathoms and feet
Survey by U.S.S. Bushnell - 1939
Compiled in U.S. Hydrographic Office
From single lens aerial photographs
It is possible that the legend doesn't mean what it says.
Perhaps you have used Photoshop to alter this image.
Or else the question of whether aerial photographs were used to map the island has been answered.
Obviously, the photos are not the sole source. The soundings were not take by aerial reconnaissance.
But it looks as though the whole process was kind of regularized.
I'm especially impressed by the notation that reads "from single lens aerial photographs."
That suggests that there may have been cameras with multiple lenses.
And, yes, that does seem to have been the case, FWIW, in 1939:
Digital Oblique Aerial Cameras"There is nothing new about taking oblique images – they have already been in use for over a century for military survey and large-scale mapping projects. Around the year 1900, Scheimflug developed a multiple-lens camera viewing oblique in 8 directions. During World War 1, the US developed a tri-lens camera. In the interwar period, engineers employed by Sherman Fairchild extended this multiple lens system to the five-lens T3A, which remained the precision-mapping camera of the US Army until 1940. The T3A can be considered as the forerunner of today’s Maltese cross digital oblique cameras as it acquired five negatives sized 5.5 by 6 inches simultaneously (Figure 2). The central lens pointed vertically, i.e. in nadir direction, and the other four, which were spaced at 90 degrees intervals around the central lens, were tilted 43 degrees away from the vertical. During a mapping conference held in Washington in 1940, the military use of the T3A was abandoned in favour of the tri-metrogon, a cluster of three K-17 wide-angle reconnaissance cameras; one pointing in the vertical and the other two at a tilt angle of 60 degrees on each side to provide horizon-to-horizon coverage. The digital variant of the tri-camera configuration has also become popular in recent years. Figure 3 shows the so-called Fan configuration when one camera is looking into the nadir, the second to the left and the third to the right. In general, the Fan consists of two or more digital cameras which have been assembled such that their optical axes are in the same vertical plane, but each camera views at a different angle resulting in a panoramic view across track. Multiple camera heads can also be mounted in a block such that they allow extensive ground coverage – equal in all directions – during one exposure. Another method to obtain oblique views is by sweeping one or more cameras across track. The scan motion allows a large field of view across the flight direction and provides oblique views. Vision Map’s A3 dual-camera system operates according to this sweeping principle."
And all of this is, of course, moot. I don't see that it gives any leverage for deciding one way or the other about the Niku Hypothesis.
Just fun facts.