Today I started measuring light intensity from this modified fixture, using a PAR meter borrowed from Tom Barr. It is a very revealing test!
First, I wanted to resolve my doubts bout the effectiveness of the simple reflectors I used. To do this I moved the fixture off the tank and put it on a simple stand so I could measure PAR at various locations and distances from the bulbs. The set up is:
I measured PAR at 11 different locations over the tank footprint area, for 8 different distances from the bulbs. Then I removed the reflectors and measured the PAR at the same locations at 2 different distances from the bulbs. I plotted that data for the location right under the center of the fixture and that looks like:
You can see that I get about twice the light intensity with the reflectors as I do without the reflectors. To me this proves that simple, mylar covered reflectors like this can be extremely effective, especially for those of us with a limited budget for lighting.