Every lens vignette, only the central portion of the projected image cast by the lens is pictorially useful. This central portion is called the circle of good definition. For each format size, there is a focal length of choice we label as “normal”. The vignette of a “normal” is unlikely to be concerning. This will be a focal length approximately equal to the corner-to-corner measure (diagonal) of the format. For the full frame 24mm by 36mm, will be 43.3mm. By tradition this value is rounded up to 50mm.
c
The APS-C (Advanced Photo System – Classic) measures about 16mm by 24mm. The diagonal measure of this rectangle and the “normal” focal length is approximately 30mm.
The difference between the two formats: The APS-C is 66% of the size of a full frame. The full frame is 1.5X (150%) larger.
Now the circle of good definition of a 50mm focal lens is adequate to cover a full frame and more than adequate to cover the APS-C format.
As to your reasoning point #1 – it is moot!
As to your reasoning point #2 – The spillover of a 50mm focal length lens is far greater than the area of both formats. Spillover must be dealt with. If not, the internal parts of the camera’s light path will likely reflect stray imaging rays. These would bathe the image recording area during the exposure. This stray light induces flare. Flare is devastating as it lowers the natural contrast of the projected image. To deal with stray light, the interior of camera is painted flat black to deaden reflections. Plus, the interior walls have baffles that mitigate reflections.
As to your reasoning point #3 – Spillover is always present in every optical system. The loss of this light energy in itself is moot except, if not controlled it is the source of flare which would induce a loss of contrast if not controlled.
As to image brightness - Any lens operating at the same f-number as another, delivers an image of equal brightness (within reason for pictorial usage). In other words the f-number (focal ratio) is valid regardless of focal length and working aperture diameter. Differences in the transparencies of the lens elements and surface reflection loss plays a part. For critical work we then resort to T-stops which are calibrated via an actual light intensity measurement.