Maybe, but the tyranny of numbers still exists. You would need enough subjects in each arm of the trial to be statistically significant. Unless there is a really strong natural history database for each disease you would need a placebo arm for each disease. I'm not sure that at this time it would be much of an advantage to do that. In the future I suspect that will change.
One of the problems that has been discussed in detail on this board is the difficulty of clear, verifiable diagnostic criteria for CNS diseases, especially in the early stages. That problem works against multi-disease trials.
Market Data 
Markets 
AI
