Include Harvested data in suggestions

There are cases where summarization excludes some harvested data, for example, the license score min-bar is 80 for ScanCode. While this makes sense for automated cases, if the user is looking at fixes, it would be good for them to see that, for example, one of the tools was 75% sure there was an Apache license. See #383 for an example of a related scenario.

So the suggestion (pun!) here is for "getSuggestions" to recompute the definition but with fuzzier settings. Then anything that is different becomes a suggestion. This should actually be pretty simple, just would require passing the fuzzy factors into the summarizer.

@storrisi there might be some changes for how the UI presents suggestions. It likely should be general as in the future there will be many different "sources" of suggestion.

Start at getSuggestions and trace how it calls the summarizer; the issue says fuzzy factors must be passed into that summarization. Review issue #383 for the related scenario and determine how the service represents suggestions. Done means harvested data excluded by strict settings is reconsidered with fuzzier settings and surfaced when its definition differs.

Written by the indexing model from the issue text.