This is fundamentally about transforming sociofugal space to sociopetal spaces. Intersections are inevitable in street networks. In fact, when designed right, they are amplifiers of value around them. After all, the intersection is the atom of the city. Or alternatively, they can be subtractors of value as they create disorienting, unsafe experience that repels people and investment rather than attracting both.
A couple points of note. An intersection is by nature a convergence point that concentrates activity. Furthermore, a city by nature is about social and economic exchange, which is friction. However, in urban areas, where we want to maximize activity in proportion to the infrastructure that facilitates it, we can't design for free flow and fast moving traffic. This is the fire hose. It erodes friction.
Today's example for the bad intersection catablog is my favorite/least favorite. The clusterf of a centerfuge where Canton, Exposition, Main, and Commerce somehow come together and nobody comes out of the black hole formed by it alive.
I've used this intersection countless times in my 12 years in the area as a driver, a passenger, a bicyclist, and a pedestrian and I never, ever know on what side I'll be coming out of this intersection. Or what road for that matter. It actually looks quite logical in plan in the way that most oversimplified suburbanized intersections look in plan. However, like this kind of engineering in general, it works for worse in practice than it does in theory. One ways turn into two-ways, road names change, flying right turns facilitate high speed movements and scare pedestrians away, while visibility and overall intuition in how to move about in it is poor. I'm not sure there is a metric that this succeeds on...no wait! There is one.
The traffic counts are incredibly, shockingly low on all sides. Listed above are the rough counts/capacities the roads are built for. In our race to avoid the bogeyman of congestion, there is something worse: traffic counts that are too low if not downright negligible. The result is economic activity can't occur, businesses, close, and thus it can't support the wait of the infrastructure built to facilitate it.
Places for cars first become invaded, then they become abandoned. On a long enough time line the survival rate of all things drops to zero. Suburbanized logic tends to expedite that time line as it externalized complexity and similar elements that can't effectively be quantified, like quality of place, comfort, and thus the impact on real estate value, which is in effect, a measure of demand and the socio-economic potential of a place.
However, those concerns are quickly erased when you prioritize the firehose for the sake of free and easy traffic flow, which is always a myth except for on rural roads. Damn those human needs for socio-economic exchange getting in the way of the engineered ideal that Detroit and South Dallas have achieved, fully ruralized urban places. Or as everybody else calls them: failed places.
However, if we were to re-work this network into a re-urbanized grid, we're able to recapture much of this leftover space that has been decorated but not activated as well as the adjacent vacant lands that are a reaction to the repellent nature of traffic flow. Density is energy condensed into a slow vibration. Compare the atomic energy of pedestrian oriented places to high speed car flow. This is the difference between a diamond and methane gas. People are attracted to one and repelled by the other.
When you successfully focus these convergence points into pedestrianized spaces, you can create great spaces that attract development and thus reposition all of the surrounding underdevelopment. When the suburbanized street network based on ruralized logic is the reason for the evacuation of investment, the network must be re-worked. There is no tactical solution outside of tactical drone strike that can fix it. Rather, it takes a fundamental reworking of the street block structure. A re-urbanization in order to re-colonize abandoned places.