>Using the same rocket ten times instead of once is an order of magnitude improvement.
Reusability is not a panacea. Space Shuttle was reusable but it was not cost-effective and there were no cost savings. Also, reliability suffered and maintenance became prohibitive. In fact, I think the consensus is that the entire program was a mistake.
>Why shouldn't it be possible to do even better than that?
Because reusability has to contend with factors like:
1) Turn around time. How quickly can you refurbish a rocket.
2) Maintenance costs. How expensive is it to recover and refurbish the rocket.
3) How many times can you reuse a rocket.
4) Failure rates. Does reusing a rocket compromise quality.
5) Decreased revenue per launch. Reusable rockets need to carry spare fuel meaning, that your cargo space is cut (and cargo is revenue), and your costs are higher, because you have to still 'pay' to ship that spare fuel along with cargo.
Factors like those may make it that even if there are cost-savings, that those cost savings are not orders of magnitude lower, but maybe something on the order of 10%-50% - which is nice, and you can disrupt the market and make a good business out of it, but it's not anywhere close to the claims that are made. And again, SpaceX has not shown in practice that they can get ANY savings from reusability.
