NetarchiveResourceStore duplicates ArcRecord functionality

NetarchiveResourceStore has to parse http headers istelf because ArcRecord won't do it for us. It appears that this is because NetarchiveResourceStore has
metadata.put(ARCRecordMetaData.URL_FIELD_KEY
captureSearchResult.getUrlKey());
instead of
metadata.put(ARCRecordMetaData.URL_FIELD_KEY
captureSearchResult.getOriginalUrl());
The former has no http:// so ArcRecord won't parse http headers for it. With a
bit of work we should therefore be able to create a much simpler and more robust NetarchiveResourceStore using API calls and eliminating our own header-parsing.