swarman
Hi Ron,
Can you please tell me whats wrong if we find probability by doing: C OR M-(C&M)??

Hey Swarman, great question! A lot of people just plug in the formula and figure it will yield the right answer, but in this case you'd likely go directly to the trap answer C (without passing go or collecting 200$)
Let's look at the formula on a simple case. 2 fair coins, and we want to know if we'll get at least one Heads. We'd use the formula to say that the chance of getting heads on the first flip is 0.5, heads on the second is 0.5, and heads on both is 0.5x0.5 or 0.25. This gives us 0.5 + 0.5 - 0.25 = 0.75, which is correct. However, the question being asked here is "at least one head", which means HH is being counted in both calculations of 0.5. That's why we need to remove it afterwards, because it's being double counted, and should only be counted once.
In this example, the question is asking for exactly one or the other, which means both should be discarded. In the above coinflip example, there would be four possible outcomes: HH, HT, TH and TT. The question of at least one heads allows for HH, HT or TH. The question of exactly one heads allows for HT or TH, but not HH or TT. The question stem can dramatically change the question and send even the most experienced test taker towards a trap.
If you wanted to solve this by a formula, the same logic would apply, except it would be C or M - 2x(C&M). Logically, HH is being counted twice, when it should be counted zero times, so you must remove both instances of it from the equation. You can use this formula if you want, but the underlying logic is probably more helpful when faced with a tricky question on test day.
Hope this helps!
-Ron