That was kind of the point with the intensity interferometer. Tens of arcminutes resolution was fine. Thus, they could use the segmented mirrors you see in the image long before the active control used in the Keck. The intensity interferometer is much less sensitive than the Michelson interferometer in terms of collecting area which is why they used such large mirrors to work on bright stars. But, it does not need the extreme tollerances of a Michelson. Much more importantly, recorded signals can be used to to produce the interference and so the senstitivity of the intensity interferometer (in terms of the amount of collecting area) exceeds that of the Michelson when there are approximately 100 or so stations. This is because the number of beams that can be combined in a Michelson is limited to less than ten or so, while all possible baselines can be measured with the intensity interferometer. This crossover figure includes some methods I've developed to improve the sensitivity of the original Hanbury-Brown and Twiss approach on a single baseline. One of the truly fundemental next steps in astronomy is to measure the rotational orientation of stars. An intensity interferometer array is particularly well suited to this since it has both spatial and Fourrier transform capability so that it is possible to measure the changing doppler shift of a spectral line across the face of a star. If you want to do work that will be included in the introductory text books, this would be it. The science that can be accomplished from an archeology of angualar momentum in clusters of stars is breathtaking, yielding perhaps the most intimate details of star and plant formation that are possible to discover.

Dish solar collectors that are built to achive fairly high temperture are essentially perfect intensity interferometer base stations for night time use and only slight additions are needed to instrument them for this purpose which can be configured not to cause any difficulties with the main goal of collecting solar energy. By contrast, configuring arrays of solar panels for all sky monitoring and intensity interferometry may require intrusive modification of the electonic design and could be much less feasible. Night time use of panels may rather concentrate on transient detection and perhaps study of ultrahigh energy cosmic rays.

If you have access to a library that has a copy of Hanbury-Brown's book on the intensity interferometer I recommend it as one of the best written books in science.

Chris