Refactor noaa and nyc_taxis tracks to include more aggregations

### Description

We need a new `aggs` challenge under the `nyc_taxis` track to test performance of the following aggregations:

* composite
* sampler
* percentiles
* derivative
* multi-terms

We also need to update the `aggs` challenge under the `noaa` track to test performance of the following aggregations:
* top-hits
* top-metrics

NOTE: `top-hits` and `top-metrics` aggregations are already part of an existing `top_metrics` challenge under the `noaa` track. We will just re-use those into the `noaa` track.

These are aggregations which are commonly used in observability-related scenarios and by products like APM.