It doesn't appear that the atom is 1 pixel, but is in fact quite a bit larger than this: there are lots of elements in the image which are significantly smaller than it. So what gives?
I used to work on trapped ion experiments, and the limitation of the atom size was always the diffraction limit, which is limited by the NA of the lens (f-stop in camera terms) and the wavelength of light. In this case, the optical system (I'm guessing a camera lens) is designed for multiple wavelengths, so it might not reach the diffraction limit at the emission wavelength, which is like 400nm. In that case, the limit would be the aberration of the camera lens, which can be wavelength-dependent. Most camera lenses aren't designed for 400nm light, which is marginally visible.
