Updates to index_points_to_lines

[dblodgett-usgs]

The function needs to be broken into a few sub components and made more flexible.

1. If the flowline match is known up front but the location along the line(s) is not known, we can pre-filter before running matcher. Could also parallelize in this case?
2. Need to support a non-aggregated case.
3. Generally need to refactor and simplify things. There's a lot of validation up front that is confusing things.

[dblodgett-usgs]

Should switch this over to using data.table and generally aim to reduce the memory footprint of the function.

[dblodgett-usgs]

Still need to refactor and support a non-aggregate use case. But much of this issue is done now.