There are a few studies at NASA (and I don't doubt, under the auspices of RKA and ESA) that bring up designs for tying Sabatier and liquefaction equipment that potentially can be used in LEO and elsewhere:
- A Liquefier for Mars Surface Propellant Production. Salerno, Lou J.; Helvensteijn, B. P. M.; Kittel, P. 1999.
- Cryogenic Fluid Management Technology for Moon and Mars Missions. Doherty, Michael P.; Gaby, Joseph D.; Salerno, Louis J.; Sutherlin, Steven G. 2010.
- In-Space Propellant Production Using Water. Notardonato, William; Johnson, Wesley; Swanger, Adam; McQuade, William. 2012.
Doherty et al., 2010:
AES funding will be used primarily to develop the experiment package that will include interfaces with the ISS
power, data, cooling and GH 2 sources, the GH 2 liquefaction heat exchangers and J-T valves, the cryogenic
refrigerator (20 W at 20 K and 40W at 90K), the cryogenic fluid management payload (mixing pump, spray bar,
liquid acquisition device) , a small (<200 liter) liquid hydrogen tank, and the passive thermal control systems. The
demonstration project will be a collaboration between multiple NASA centers, including Kennedy Space Center
(KSC), Johnson Space Center (JSC), Glenn Research Center (GRC), and Goddard Space Flight Center (GSFC). As
envisioned, KSC will responsible for the design and fabrication of the experiment pallet, the passive thermal control
system, the fluid components, and the pallet integration and systems engineering; JSC will be responsible for
integration of the pallet to the ISS, the command and control system, and the orbital test operations; GRC will be
responsible for design, fabrication and integrating of the CFM components into the storage tank and the system
modeling; and GSFC will be responsible for the development of the 20W at 20K cryogenic refrigerator.
Why hasn't this been done before?
The simplest answer is that you need a bunch of liquefying machinery (or compressors) to transform methane into usable state, and that machinery was not fitted into the original Russian modules (which do the thrusting).
Another consideration is that storable hypergolic propellants are much more studied at the time.
Methane-oxygen engines are very much in vogue nowadays, with Mars exploration and "clean" rocketry, but that was not always the case.
A related question: Why does the International Space Station produce so much methane?