Repo-analytics was a useful web app that was similar to this one.
It would graph the view history of github repos, but it also tracked the clone history.
More importantly, to gather longer-term data and get around github's meager 2-week data timespan, the backend server would regularly check each repo's analytics and get new data.
For a while there I had over two year's worth of data, and it was really useful to see every spike in clones and cross-reference that day to a news article or YouTube video referencing the project in question.
But then repo-analytics went down, and I believe it is never to return.
Suggestion:
track git clones
store viewership and clone data long-term and regularly collect new data
graph the clones and views with a graph that can show all-time, or can be made more specific to only show the past month, for example.
Keep a running total number of views and total number of clones somewhere. This feature never existed in repo-analytics, so back when it worked I tried to reverse-engineer the y-values of every point on the graph and correspond them to a number of views/clones. That was a mess.