A quick search of the forums indicated that several users have eliminated the problem by going to a split design. This made no sense to me and would be a pain to accomplish with my panel due to the location of my electronics, so I dug around the internet to find a solution. I didn't find a solution, but I did find a find a better explanation of the cause: My original polarizers probably had a compensation film, and my new ones don't have it.
This link explains the purpose for compensation films: http://www.lci.kent.edu/boslab/projects/op...tion/index.html
In a nutshell: Light leaving the LCD's TN cell will only be correctly polarized relative to the analyzer (the exit side polarizer) if it passes through perpendicular to the panel. Adding a compensation film serves to allow light passing through at an angle to be correctly polarized.
Using the split configuration, the majority of the light will be parallel to the light path and no compensation is needed. This is why going to split optics fixes the problem.
Let's be clear: If you have an unaltered LCD, you can likely use it with no problems at all - split, unsplit, whatever. Because of the compensation film, light can pass through at an angle and it will still work. Replacement polarizers will probably not have compensation film and will not work with off-axis light. (This is unrelated to the inherent inefficiency of the panel with off-axis light. That problem is caused by the internal structure of the LCD blocking light that passes through at an angle. Picture what happens when you look through open vertical blinds and then walk to one side or the other. )
Warning to all: If you have removed your polarizer, you have likely removed your compensation film and will be limited to split optics. That is, unless you can find replacement polarizers with an appropriate compensation film.
I hope this helps someone else not waste as much time as I have on this issue. I guess I'll have to figure out how to go split with my projector.
