Add better indication of what scores mean

[keyserj]

Describe your issue

It's not easy to tell what node and edge scores mean. Having docs/tutorials help, but this feels like something that can be conveyed within the app, which would be better so that people don't have to go searching.

Note: this issue is independent of what the scores actually mean - whatever they mean, that should be understandable in the app. (related: #452)

As of this writing, node scores indicate the extent that "this node is an important [node type]", and edge scores indicate the extent that "node 1 [edge label as a verb] node 2" (e.g. extent that node 1 causes node 2) (note: scores 1-4 indicate "does not", 6-9 indicate "does").

Solution you'd like

Potentially when the score select/compare pie is shown (hover/click), a text explanation can appear next to it.

Questions:

• what should the text explanation be?
  • probably similar to the root claim text. would be nice if the root claim node could just show, but the text usually doesn't fit well in the root claim node! maybe the text should be modified slightly to be explicit that the score means the extent that this text is true?

Other thoughts:

• potentially the top claims can also show next to the text explanation, but that may be a lot for the UX to do on-hover. maybe there could be a quick details-like button to make these arguments show?
• I think with this we should remove scores from Justification edges, because Supports/Critiques are not intended to be standalone, and so their node scores are based on how well they support/critique the parent claim. Which leaves no difference between the node and edge scores. This change will help clarify that. See Discord thread for context.
• this UX could be good:
  
• I think score shape could be really helpful for distinguishing the meaning of the scores. e.g. square for importance (nodes), diamond for truth (edges), circle for average, octagon for disagreement

Alternatives you've considered

No response

Additional context

• suggestion from this discord thread
• is it useful for edge scores to mean "probability that A causes B" sometimes? should they be a combination of extent & probability?

Technical ideas and questions

• re: score shapes: seems like clip-path would be good for these, see https://css-shape.com/ for shapes using that, https://bennettfeely.com/clippy/ for creating paths for it
  • annoyingly borders get clipped with this method too, probably would need to do something like this https://stackoverflow.com/questions/31854185/how-to-add-border-in-my-clip-path-polygon-css-style