Actually Tesla3, ran some numbers that showed exactly this. He dismissed it as "inconclusive".

I measured the width and the height of the oil rig in the two photos.

I attributed an error to all of my measutrements of +/- one pixel.

Within experimental error the height:width ratio was the same in both pictures (see below).

So from a proper, scientific viewpoint there was no recovery of height with magnification.

If anyone wants to ignore the experimental error, please go ahead.

But that would be un-scientific, anti-scientific or pseudo-scientific.

----------------------------------

Zoomed out Height (H1) = (30 +/- 1) mm (where 1 mm is the width of one pixel)

Width (W1) = (58 +/- 1) mm

Fractional error in H1 = 1/30 = 3.33%

Fractional error in W1 = 1/58 = 1.72%

H1/W1 = 0.517

Fractional error in H1/W1 (FE1) = 3.33% + 1.72% = 5.06%

Zoomed inHeight (H2) = (83 +/- 1) mm

Width (W2) = (155 +/- 1) mm

Fractional error in H2 = 1/83 = 1.20%

Fractional error in W2 = 1/155 = 0.65%

H2/W2 = 0.535

Fractional error in H2/W2 (FE2) = 1.20% + 0.65% = 1.85%

Comparing the two pictures(H2/W2)/(H1/W1) = 0.535 / 0.517 = 1.0353

Fractional error in (H2/W2)/(H1/W1) = FE1 + FE2 = 5.06% + 1.85% = 6.91%

Actual error in (H2/W2)/(H1/W1) = 6.91% x 1.035 = 0.0715

ConclusionThe shape of the rig in the two images is the same within experimental error.

(Given that: 1.0353 - 0.0715 = 0.9638; and 0.9638 < 1.0000)