It is difficult to answer with certainty about which camera you need, especially since you provide no details about your microscope, your future specimens and what experiments shall be done. So I will try some general suggestions.
As you probably know, fluorescence units are arbitrary, not absolute. Fluorescence intensity depends on several factors: chiefly on the quantum yield of the observed specimen and the intensity of the excitation - in addition to the obvious dependence on wavelength. The microscope optics, including the objectives, dichroic mirror and excitation and emission filters also play a role. Often one needs to work in a dark environment to avoid stray light. So a given camera might be suitable or not.
I think that low light level performance (sensitivity, signal-to-noise) of the camera is crucial, and is more important than pixel resolution. Full control of the camera by means of the computer software is, I think, also very important for research. A dedicated computer card like the PCI you mentioned, is IMO only necessary if you plan a very high rate of transfer of high-resolution images to the computer. Otherwise, the dedicated card on obsolete computer motherboard is not critical.
I am guessing that you are starting a research on a budget. Top level fluorescence research data rely on dedicated microscope cameras - say, Olympus DP-74 (just an example) or, for confocal microscopy, photomultipliers. The latter are much more sensitive and fast and linear than cameras (a generalization, not overwhelming). And the microscope manufacturer's software. I would choose such camera rather than a DSLR.
One other point that you might want to consider is the coverage of the field of view. For example, the Olympus DP-X (X=70, 73, etc) camera, on top of a BX-series microscope is a wonderful modern setup. Yet, FOV coverage is far from complete, unless you specifically add a relay 0.3X or 0.5X lens above the objective.
May I suggest the following approach: prepare some test slides. The easiest, and most reproducible, are fluorescent micro- or nano- spheres. They are made of latex or polystyrene or other polymer, at sizes of 100nm-10micron (at least), and are stained with a monochromatic or polychromatic stains. They give off a very bright fluorescence that is quite immune to photo-bleaching. They are uniform in size. Can be mounted dry, or in a liquid or resin medium. They are available from most chemical/biological supply houses. Some photos of such slides, perhaps with a (loaned) camera, will provide a qualitative and quantitative estimate of the performance of the camera.
Zeiss Standard GFL+Canon EOS-M10, Olympus VMZ stereo