Dear Kathryn,
     there is no reason not to use a wide-field system for polarization 
anisotropy imaging. I am trying to address some of your concerns:

- reflected light could keep its polarization at a large extent, but fluorescence 
cannot. Therefore, with the use of one photon excitation, the maximum 
anisotropy one can measure is 0.4 not 1, i.e. you will always detect signal
- the usefulness of polarization anisotropy is expressed when you make a 
ratiometric measurement; if you just measure the light passing through an 
orthogonal analyzer one can never know if the fluorophore is at higher 
concentrations or if more depolarized. If this was done on a plate reader, I 
could infer they were using an homogeneous assay where fluorophore 
concentration is fixed
- multi-photon systems provide an advantage in anisotropy imaging because 
the fluorescence emitted by your sample can exhibit much higher anisotropies 
in absence of rotation and energy transfer (0.57 vs 0.4). Furthermore, a multi-
photon system provide pulsed excitation and permits to perform time-resolved 
anisotropy. TR-FAIM allows for more information to be acquired and higher 
dynamic range
- I do not have references at hand for wide-field systems, although you may 
find information looking for papers by Axelrod, Jovin, Gerritsen groups or 
checking Lakowicz's book; however, the wide-field system I remember was 
published by Rizzo and Piston on Biophys J 
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1305173). If you 
need more information, I could send you an unpublished work where I describe 
a setup for anisotropy.
For a wide-field system you may use a dual-view which splits the two 
orthogonally polarized images on the CCD camera and allow one to perform 
polarization anisotropy with a single detector

I hope this is enough for now, 

Best regards, 

Alessandro Esposito
Laser Analytics Group, Cambridge University
www.quantitative-microscopy.org
www.wikiscope.org