Need help with sun simulator design

I've been doing a fair bit of solar thermal collector testing over the last couple years.  The way I do that is to wait for good steady sun, and then test the new collector design as well as a known baseline design side by side in the same sun conditions.  This works pretty well, but you have to wait for the sun, and you always have to test a baseline collector along with the new collector to account for sun variations.

To make the testing easier, I've been working on an inexpensive indoor sun simulator.  This gadget uses a few high wattage metal halide lamps (like street lights) to reproduce lighting levels similar to the sun on an up to 2 ft by 4 ft test collector.  This setup is looking pretty good, but I'm having trouble coming up with a reflector design  that puts enough of each lamps light output onto the collector.

So, if you have a minute and want to have a look at what the current sun simulator/reflector look like and pass on any ideas for improvements, I'd appreciate the help.

Details on the sun simulator and reflector ...

Gary