Ability to aggregate (sum) over time range for point data in order to map Covid-19 data

**Is your feature request related to a problem? Please describe.**
I've been looking into using Kepler.gl for the typical Covid-19 data: daily case counts, aggregated per region (Lat/Lon coordinate of centroid), with a time-based filter to visualize development over time.
The problem I ran into is: with daily data on a fixed set of coordinates, using simple Points (circles) per Lat/Lon, the only sensible time range is 1 day. Increasing the time range results in Points being drawn on top of each other, which makes no sense in this context.

Example map: https://jvlanalytics.nl/assets/blog/2020-05-01-evaluation-keplergl-covid-19-data/covid-19-map.html

Furthermore, the 1 day time range is not selectable (it's too small), so the only time control I have now is Play/Pause.

**Describe the solution you'd like**
Ability to aggregate data for Points (e.g. sum, mean), *if* the Points are on the exact same coordinate.

In the map I linked above, I'd set aggregation to `sum` and I'd expect the Point radius to increase as more days are included in the time range (sidenote: the map I linked has cumulative cases; I'd instead use daily new cases if this were possible).

It would also be great if a relatively small time range, like in my example, would actually be selectable in order to scrub through the timeline.

**Describe alternatives you've considered**
The map I linked already has my workaround: using cumulative cases rather than daily new cases, sticking to a 1 day time range and informing users not to change the size of the time range.

**Additional context**
I ran into this issue while working on this blog post, which has a bit more details:
https://jvlanalytics.nl/evaluation-keplergl-covid-19-data
See heading "Map 1: Circles like the famous John Hopkins Covid-19 map"

This kind of aggregation behaviour would also be great for Polygons.