I agree, it depends on the photo.
Apart from this specific foto tho, i say the bokeh does depend on the viewing distance.
Here an example (not claiming its a good photo, just something i had at hand). If we look at the first image (a crop), we see the bokeh looks decently creamy. But if we look at the second image (the full image), the background seems distracting and not creamy.
Your example doesn’t really seem to be about viewing distance, though.
Cropping does affect DoF, but the underlying assumption is that the original and cropped images are being viewed at the same size.
At a basic level, DoF can be reduced to two factors – physical aperture and magnification. DoF is inversely proportional to both, i.e., a wider aperture and more magnification both mean less DoF. Magnification is determined by subject distance, focal length, image enlargement, and viewing distance. Thus, if you crop an image but view it at the same size as the original, you’ve relatively increased the magnification of the cropped image, thus decreased the DoF.
Conversely, if you reduce the magnification, you increase the DoF. Thus, if you downsample an image or view it from further away, the DoF is increased.
Side note that DoF calculators assume a standard output size and viewing distance, typically an 8x10” print viewed at ‘arms length’.
Side note 2: it’s a commonly held belief that smaller sensors have deeper DoF. That’s incorrect, they have shallower DoF for the reasons described above. However, the context of that belief is typically comparing the same focal length and framing, and when doing so you must be further from your subject with the crop sensor. Further away means less magnification means more DoF.