If that's your measurement situation, sure. Generally I'm providing at least one laser beam, so can usually know where I am, more or less. Sometimes I need to know the absolute photocurrent, in which case I usually use a slow LPF on the other end of the diode where it doesn't get in the way.
For really wide range things I like to use diodes as feedback elements (diode-connected transistors really, of course). Diodes have noise temperatures of half the ambient (150K at room temperature), and if you use a bunch of them in series, the gain and load resistance adjust themselves so you can stay in the shot noise limit over a much wider dynamic range than you can with resistors. Another nice approach is just to stuff a photodiode into the input of a nice quiet UHF amp--Miteq sells some that get down to 25K noise temperature at 300K ambient, and
10K at 77K ambient.I actually just ordered some BF862s and OPA657s to try out--with all the device work I've been doing, I haven't designed a front end in ages, and I need to do a few for the second edition anyway. I don't use anything like the same variety of parts as you do in AoE, but still, I can't very well go on singing the praises of an op amp that went out of production in 2004 (the LF357).
(I just got back a few sample chips that I hope will be the first real photovoltaic antenna-coupled tunnel junctions. Gold antennas and yttrium / nickel oxide / platinum TJs. Odds of them working: probably 30%.)
Cheers,
Phil Hobbs