The timing option is about when you make delivery and it's more or less worthless. The optimal delivery date, for the most part, boils down to whether or not carry for the cheapest-to-deliver (CTD) is positive. If carry is negative, early delivery is usually optimal. If carry is positive, late delivery is recommended. The edge case is when the switch option (more on this below) is very valuable but carry is negative. Then there's a tug-of-war. If switch option is not valuable enough to offset the negative carry, you'd still deliver early. But if the switch option is very valuable, you might want to deliver later in spite of the negative carry (as soon as you make delivery, you forfeit the switch option completely, assuming it has any value.)
The switch option is about which bond you deliver.
- Generally speaking / in most models, the switch option (also known as quality option) is about the possibility that the cheapest-to-deliver (CTD) can change between now and the delivery date, where the delivery date is pre-specified to be either the first delivery date or last delivery date, depending on the carry profile (this is not as bad as it sounds, because the timing option, as mentioned above, is pretty much worthless).
- More complex models might account for the end-of-month option, which is really another switch option (as opposed to being a timing option). Remember that futures price stops changing after the last trade date, but cash bond prices still do. So between the last trade date and the last delivery date, it's possible that the cash market moves by enough that the CTD can change again, after futures have stopped trading. The EOM option is usually worthless too, but for the classic bond contract, it has historically been worth as much 3-5 ticks.
Edited to remove incorrect characterization of the wildcard option. Please refer to @dm63's excellent discussion in the comments below.