The way that we design quiet residential streets could be more efficient. The modified grid model is somewhat waseful--in terms of space, asphalt, and energy.

A honeycomb approach--where each 'block' is an elongated hexagon--could handle the same traffic, with a slightly smaller paved surface area. The ratio of living space to car space would be slightly higher.

The small triangles at the end of each 'block' could be planted with some attractive vegetation, to enhance property values--and in areas with sufficient rainfall for deciduous trees, to partially block the Summer sun from heating up the houses on East, South, and West ends of blocks. On the North-pointing triangles, we could plant attractive, low-maintenance shrubbery.

The traffic intersections would be 3-way, rather than 4-way. With the greater visibility at 3-way intersections, if all motorists remembered the yield-to-the-driver-on-the-right rule, we could simply slow down at some intersections, rather than coming to a complete stop--or even a California stop. This would be especially true for Northbound cars on 'blocks' that have a NS long axis. We could get by with fewer stop signs. Neighborhood trips would be a little faster. Autombile fuel efficiency would be a little better. And air pollution would be a little less.

Caveats. Obviously, we're not going to uproot existing houses to make honeycomb neighborhoods. The idea would mainly apply to new housing tracts. Moreover the topographies of some residential areas won't lend themselves to a honeycomb design, just as they don't for a traditional grid design. The major traffic arteries should continue to be designed in the same way.

If you're looking for a buzzphrase to summarize the primary virtues of the honeycomb residential street model, it's perimeter efficiency.