"An exposure lasting several days is by definition several shorter exposures, one per orbit. The real problem with long exposures is synching up all of the various exposures. If the telescope is not pointed in the exact same direction for each exposure, you loose that. Remember, the Earth and Hubble's orbit relative to the Earth move each orbit. "<br /><br />Still not sure I see your point. With detector noise you generally can't observe very long any way, so you have to do short exposures. You also want a little "wiggle" so that you hit different sample diodes on different exposures to remove detector effects, especially on such sharp point spread functions as HST has. (Though there are exceptions depending on the science.) It is trivial to add up numerous images of say the WFPC with the pointing accuracy of HST. If HST was out at L4-L5 I don't think you would see any significant improvement in stability. I am not diagreeing, just don't see it as really that big of an issue. Now, being out a place like that really does open up some nice options of field of view. It will be awesome when Kepler gets up there and stairs at the same point in space continuously. Man, will we get a lot of great data from that. Who cares if we actually find a planet, I'll take all the data they throw away!