How may one question the general acceptance that sensory input first enters sensory memory before being transferred to short term and then long term storage?