add clearer distinction between purpose of sections 4 and 5

[nickreich]

• Section 4 is titled "Demonstration of functionality" and uses a minimal subset of "example" data from the FluSight project to demonstrate basic functionality.
• Section 5 is titled "Case study: Weekly incident flu hospitalizations" and uses a very complete set of forecast data from FluSight over several seasons.

Nonetheless, the start of section 5 reads: "To demonstrate the utility of the hubEnsembles package..." and then goes on to describe the FluSight data, which looks very similar to what was used in section 4.

My suggestion is to make the distinctions between the purposes of these sections and the extensiveness of the data in each more clear to the reader. Some specific suggestions

• Introduce FluSight in 3.1 or 4.1 and say clearly that some toy example data will be used for basic functionality in section 4 and for showcasing a more extensive analysis in section 5.
• Consider changing section titles to "Basic demonstration of functionality" and "Example in-depth analysis of forecast data" or something like that
• Mention at the start of both sections 4 and 5 the distinctions between the data and analyses presented in each section.
• State more clearly at the start of section 5 that the purpose of this analysis is to show a "complete" analysis that compares and evaluates ensemble model performance, with the goal being to choose a "best" ensemble approach to use in practice. This motivation and use-case is not clearly stated currently in the start of section 5.

[lshandross]

The earliest use of the flusight data in the manuscript will be replaced by example forecasts data from hubExamples based on the discussion on Issue #42. This way, all earlier examples will be using data from hubExamples (Issue #43), with the package being introduced early in the manuscript.