Review feedback from #4

[sanket1729]

Based on @apoelstra's comment https://github.com/sanket1729/elements-miniscript/pull/4#issuecomment-788104864

• the nits that I listed (Addressed the nits are as_cov returning result instead of panic)
• need to investigate the fuzz failure (fixed in Fix fuzz test #5 )
• should change the itempref stuff to allow a variable number of CATs based on the length of the prefix(Unaddressed, see https://github.com/sanket1729/elements-miniscript/pull/4#issuecomment-788411479)
• I guess we should fix the secp context creation in the interpreter. I have a few thoughts about these.
  It's not a super high priority. (postponed, created Think about to interpreter API design for checksigfromstack #6 as a reminder)

[apoelstra]

This took me a bit of thinking :) I think that max_elems is misleadingly named because it's actually one less than the maximum number of elements. I think it'd be better to leave the - 1 in place (no parentheses are needed btw) and to change

        let max_elems = MAX_SCRIPT_ELEMENT_SIZE / MAX_STANDARD_P2WSH_STACK_ITEM_SIZE;
```
to

    let max_elems = (MAX_SCRIPT_ELEMENT_SIZE + MAX_STANDARD_P2WSH_STACK_ITEM_SIZE - 1) / MAX_STANDARD_P2WSH_STACK_ITEM_SIZE;

```
or even just

        let max_elems = MAX_SCRIPT_ELEMENT_SIZE / MAX_STANDARD_P2WSH_STACK_ITEM_SIZE + 1;

since these numbers are fixed and we know that one is not a multiple of the other.

[apoelstra]

ack 15e117f