The wall profile is curved. The forms are supposed to be like this.
Perhaps I should have said "angled". Curved is not correct in the most accurate geometric shape of this form sitting in the spillway. The design of the backside or outside face of the vertical spillway sidewalls are angled outward as a buttress (thickening) outward to the base of the bottom of the sidewall. Of note, if you look at the photo of the near sidewalls closest to the camera you can see the upper horizontal line where this angled buttress starts.
The "angled buttress", where the widening starts above, is straight in its widening to the bottom or base of the poured sidewall. To date, there are no construction photographs showing a matching angle to the base of any of the backsides of the constructed sidewalls that match the "bend" location of that particular wood form laying on the spillway. The "bend" line is too low (if referencing the safety walkway as the top of the form).
That is why the discussion of a potential adhesion issue causing a bending action if a release agent wasn't working properly.
I suspect that an adhesion problem may have caused an unintended "bend" in the area that has the larger gap between these "spacers" - as this would be the area of the greatest fulcrum stress from an overly high adhesion during a "pull removal". This would match the characteristics of the form laying in the spillway.
Another confirmation & design insight of the upper "bend" line is revealed in the 2nd to last unpoured sidewall section in the mid-left of the photograph. The "white" dots are the concrete spacer blocks that meet up with the surface of the wood form. There are three rows of these spacers. This reveals the location of the upper "bend" and also reveals the rigidity necessary of the form between these "spacer sections".
Another important note: All of these large photographs have exhibited a "fish eye" lens curvature effect. This effect is pronounced such that vertical objects are "tilted". In the photo you can see a "tilt" to the power line towers that is an illusion. In the case of the interesting lower "bend" in the wood form the camera shot is "in line" with the vanishing line of the photo (i.e. parallel). Thus an accurate depiction of the geometry is revealed in the form.