I wish I could still command the math to figure this all out, but perhaps one of the other Zoners still has it at his fingertips. In general, the resolving power (the angular separation of two objects) which can be achieved by any optical system depends on two things: the wavelength of light you are observing at, and the size of the lens or mirror (or parabolic antenna, you can do this with radio, too!). For example, a binary star system will appear as one star in a telescope until the objective lens is big enough to "split" it into two. This splitting is easier at shorter wavelengths. This is not a technological problem, it is a result of the wave nature of light and the way that telescopes work. There are just limits imposed by nature.

There is no theoretical limit in how big we can build telescopes, but of course there are practical limits. One way around this is to build interferometers, which are pairs of telescopes, connected by optical or data links. These devices can be built so that the separation of the two lenses can be made large enough to simulate a huge telescope of that aperture. For example, radio telescopes on opposite sides of the earth, simultaneously observing the same object and combining their data can have the same resolution as a radio telescope the size of the earth! This has actually been an accepted radio astronomy technique now for years, and resolution in the arc-second range has been achieved in the microwave spectrum at frequecies in the centimeter range. To give you some grasp of this, a resolution of 1 arc-second at 3.26 light years, could split two light sources 1 astronomical unit apart. Astronomers will recognize this as the definition of the parsec, the distance at which the earth-sun distance subtends an angle of 1 second of are, or 1/3600 degrees on the sky. This is comparable resolution in the radio spectrum to a moderate size opticaltelescope in visual light!

In space, we could build optical interferometers operating at visual light frequencies which could be hundreds, perhaps thousands of kilometers apart. The resolutions would be phenomenal and (in theory) unlimited. Of course, there are probably practical limits due to engineering issues.
I'll dig around and try to find the appropriate equations (or perhaps some astronomer reading this can actually crank out the numbers for us).