Clarified limitations of workgroup dispatching and sizes and corrected how workload...

Clarified limitations of workgroup dispatching and sizes and corrected how workload is determined and passed into the shader, including bytes available for each pass (hint: we will always need to pass in at least some crypto bytes due to sheer size of computation, i.e. 256x256 is 65k nonces but only 2 bytes of 8 required). Start filling out actual hashing function.