During undergrad, me and some peers (all chemists) talked about having such a network of air sensors. Having good resolution on air quality measurements could drastically improve the quality of life of people living in big cities. We never knew how to monetize this, however. One way we thought about was to install a private network of sensors (inside and outside houses) and then sell the datasets to research agencies. Doesn't see too profitable though.

Ideally what you need is a machine that can take different measurements: wind speed and direction, temperature, humidity, VOC levels, ozone, nitrogen compounds and general particulate matter.

A static network of sensors is a great start, but your measurements basically define a surface. What you'd want is a fleet of drones which could "stand still" in formation, let's say a 50 x 50 x 50 drone cube each 1m apart, and take volumetric + time measurements. That would be _awesome_ :)