Answer from cs61c-ev (Roland Carlos 16583196) for Question 1
If we did not pipeline control, we would cause unpredictable results in our datapath. As one instruction is going through the datapath, another instruction may be running in parallel (although a cycle behind). However, the control signals for that new instruction will override the signals for the first instruction, so the first instruction, which may have been an innocent nop of all things, could actually end up doing addition or writing something to memory/registers. Basically, each new instruction's control signals will override the previous instruction's signals, causing the previous instructions to work improperly.