Light microscopy allowing sub-diffraction limited resolution has been among the fastest developing techniques at the interface of biology, chemistry and physics. Intriguingly no theoretical limit exists on how far the underlying measurement uncertainty can be lowered. In particular data fusion of large amounts of images can reduce the measurement error to match the resolution of structural methods like cryo-electron microscopy. Fluorescence, although reliant on a reporter molecule and therefore not the first choice to obtain ultra resolution structures, brings highly specific labeling of molecules in a large assemble to the table and inherently allows the detection of multiple colors, which enable the interrogation of multiple molecular species at the same time in the same sample. Here we discuss the problems to be solved in the coming years to aim for higher resolution and describe what polarization depletion of fluorescence at cryogenic temperatures can contribute for fluorescence imaging of biological samples like whole cells.