I'm playing around with pool share difficulties looking for significant 24 hour differences.
Do I have 63 335Mh miners or one 20Gh miner. I imagine the answer is both but how do these individuals work in unison on the same problem at the same time?
Your one bfgminer (one running instance of the program) connects to the Slush's pool as one worker, doesn't it? Then, the pool can only see this one 20Gh/s miner, I think.
I believe that the pool now sets the target difficulty for your miner dynamically according to the miner's hashrate.
The miner program behaves as one big worker in the situation described. Thus it is optimal to use appropriate difficulty setting for that 20Gh/s worker, to conserve bandwidth and server resources. There's no need to set it to 1. That might have sense for 63 individual workers, each of them connected separately to the pool, but this is not the case AFAICU.
But, as I've said above, currently, the difficulty for each worker is set dynamically.
A major wake up for me. I don't have to solve anything. Whether 1 or 63, I'm rewarded for the effort to send the correct answer even if I don't and never do. I'm paid for effort, not results. Honestly, I thought I had to solve something
You're paid for your effort. But the proof of your effort must be some results. That is, you definitely have to solve something.
What do you mean by correct answer? If you mean finding a block, then yes, there's only a little chance that you will find it in the current round. But correct answer may mean any valid share of the difficulty equal or higher than your target difficulty. Each of these shares is the proof of your effort and each of them is a product of your solution of something.