As I understand it, there's lots of problems with things like this, and that's one of the potential issues, but I would imagine that this mostly uses raised beds with clean soil brought in from elsewhere.

We thought about doing some gardening when we had a backyard in an apartment in New York, but even grass wouldn't grow in that soil - it didn't help that the soil was littered with debris like glass, concrete, random bits of metal, etc.

It's pretty cheap to have soil samples tested, but it winds up just confirming what you expect - our NYC backyard had 18x the upper limit for lead! Definitely raised beds.

Hydroponics is also perfect for polluted urban areas. Generally produces a higher yield per area and absorbs co2 at a higher rate; with the caveat being that there is potentially a higher rate of maintenance.

Does it really absorbe co2 at a higher rate once you factor in the co2 used to create the nutrients and chemicals you feed the plants?

As well as the fact that for every apartment-size space used by a hydroponic farm, that's a family that likely has to commute into the city?

Assuming you're not converting some dense residential zoned area into a garden and no one moves away due to their dislike of hydroponics, I don't think it's going to tangibly effect traffic rates which are show to fluctuate based on availability. Ie. after reaching a certain point of urban density; if you increase the availability of the road (by widening it for example), you'll increase the number of people driving such that the availability is reduced again.

If you want to reduce traffic and co2 further while increasing livable, walkable space in dense urban areas - then that can be accomplished by reducing the width of the roads.

The idea is that everything comes with an opportunity cost - do you build a garden, or do you build more apartments?

The traffic comes from the fact that every family not living in the city, but living near it, ends up commuting there anyway.

Locally, definitely. If urban heat island effects are a problem then co2 displacement would be beneficial. There is also the future possibility of processing waste streams locally and using that.