The config param `use_nextalign` has not been used since
<2839278>

This commit removes the rules `mutation_summary` and
`build_mutation_summary`, both of which called the removed script
`scripts/mutation_summary.py`.

I was looking into updating the script to work with Nextclade v3 outputs
but it seems like the rules using the script are not actively used in
the Snakemake workflow!

In Feb 2021, the script and rule `mutation_summary` were added and the
only use of the output `results/mutation_summary_{origin}.tsv.xz`
was to be uploaded it as an intermediate file of the builds.¹
In Jan 2022, the upload of the intermediate file was removed² and I
cannot find any other uses of the file so the rule `mutation_summary`
seems unused.

In March 2021, the rule `build_mutation_summary` was added to support
the rule `add_mutation_counts`.³ However, `add_mutation_counts` was
later updated to use `augur distance`⁴ which does not require the
mutation summary output. I cannot find any other uses of the output file
`results/{build_name}/mutation_summary.tsv`, so the rule
`build_mutation_summary` seems unused.

¹ <#562>
² <#814>
³ <3b23e8f>
⁴ <d24d531>

The only downstream use of the insertions and translations outputs of
`align` rule was the `mutation_summary` rule which was removed in the
previous commit.

Even though this is no longer doing translations, we are still keeping
the `annotation` config param and the `defaults/annotation.gff` file
for codon-aware nucleotide alignment.
<#1160 (comment)>

Major change includes the removal of the insertions.csv output from the
`build_align` rule. This output was only used by the
`build_mutation_summary` rule, which was removed in a previous commit.

Note: Using `nextclade` instead of `nextclade3` to allow the workflow to
continue to work with Snakemake's `--use-conda` interface.