I am trying to do some back of the napkin math on the proposed SpaceX DragonXL to figure out how much extra fuel it needs versus the original dragon, but I don’t know how much parasitic weight from extra tanks to account for.
According to Northrup Grumman, you can get a 23.3 kg tank that can hold up to 454 kg of Hydrazine. (This tank is apparently used by NASA's TDRSS satellites, see https://www.northropgrumman.com/space/diaphragm-tanks-data-sheets-sorted-by-volume/ and look up the item number 80263-1)
If we want to use this tank's stats to make this a bit overly abstract, we might be able to say every kilogram of hydrazine will require about 0.05 kg of additional tank mass. This ignores the fact that if you put more tanks on the spacecraft that will likely mean the spacecraft will need to be larger and have more plumbing, and that smaller tanks will likely have worse mass fractions, but it might provide a general idea of propellant mass versus tank mass.
To answer the question as stated in your title: it varies. A lot. But not for the reasons you seem to think--regardless of propellant type, most propellant tanks mass about the same amount. Why? Because they are generally made of the same stuff: either composite overwrapped pressure vessels (COPVs), or lightweight constructions of advanced Al-Li alloy. Even RCS systems generally feed out of the same type of tanks, with perhaps a bladder inserted to help push the fuel into the feed system without needing to ullage the tanks.
Also, even the most corrosive or hardest-to-store fuels don't need special materials. Red fuming nitric acid, ever since they figured out you could throw HF into it to corrode the tanks (get this) before the RFNA does, thus forming inhibited RFNA (IRFNA) can be stored in pretty much any metal vessel HF will corrode, which is like, all of them. And liquid hydrogen needs insulation and a special liner to help it from a) evaporating and b) slipping out between the metal atoms of the tank, but it doesn't need a special tank material.
So really, what changes tank mass? It is the thickness of the walls. This thickness is required to either support the mass of the stages above the tank, or to hold the pressure within the tank. It is this that varies tankage mass the most, as the primary factor is the pressure within the tanks. By my calculations, even very large pump-fed stages with a lot of payload above them, such as S-II, still had so much pressure in the tank that they were more like balloons than skyscrapers.
But not all tanks are that way. S-IC wasn't. And it was a lot heavier per volume of propellant than S-II, not because of its pressure within (it was about the same) nor its different propellant (not relevant), but because S-IC was made of stringer-skin-frame construction and S-II was made of gridded construction. The former involves a lot of welding, and the stresses of the tank run through the stringers & frame. The latter is milled out of a single piece of metals, and the skin of the tank is the structural component, with triangles (or rectangles) removed to save mass.
The former is a lot cheaper to do than the latter, but it is much more heavy.
I think this remains an interesting question on its own, but you might get the info you're looking for if you ask a new question specifically about the mass of the Dragon tanks. Also, I'm pretty sure that the Dracos run on MMH, not neat hydrazine.