Improve uncertainty visualization on /calc page

[hazarane]

Description

From Soeren (https://epidemicforecasting.slack.com/archives/C010KK204TG/p1593434663481500?thread_ts=1592922165.444400&cid=C010KK204TG):

• When moving the slider for R0, the red uncertainty band for R doesn't move.

• The R0 slider itself shows an uncertainty band but when the user moves the R0 slider there should be no uncertainty shown because (I believe) the calculations use a fixed value for R0, not a distribution. It seems to me that the best solution is to just not show or use uncertainty for R0 at all since we don't show country-specific R0 anyway.

• The uncertainty for R seems to be hard coded / not calculated? This would be important to fix. Here's how you can compute it: You have a fixed value of R0. For each combination of NPIs you get a distribution over its effectiveness - that's the distribution of the products of the different NPIs. I.e. the same distribution whose median you calculate. For that distribution you can get the STD, variance or whatever you use to make the uncertainty bands.

• Instead of shaded color regions with no legend, it may be better to show standard deviations and 95% CIs.

Acceptance Criteria

When the user displays epidemicforecasting.com/calc page, all the described changes are applied there.