I dont think the GP is asking for a universal reference frame in the same way physics describes it. I think he's asking for a universal programmatic reference frame to describe the offsets that sattelites and lunar stations have.

Even though time runs faster on sattelites, and the signal delay differs based on position. Because the vectors of sattelites and the moon is constant, as well as the moons gravity, you could calculate what those offsets should be and synchronise clocks anyway.

To put it another way, you normalise the clock drift and the delay, whats left over is the same clock drift you'll need to account for when you have a quartz clock in your computer and a time server that runs an atomic clock.

The point being, clock drift can become a huge issue unless you define a specification that allows any equipment to sync up with any other clock in Time And Relative Dimensions In Space.