For fun, I decided to do a rush-hour traffic count of the only intersection between me and my primary grocery. That's the corner of Baxter Ave and Goddard Ave, for my internet stalker fan-base. In that hour, I recorded 1167 crossings. 

• 95.7% were in cars (including 2 commercial trucks and one paratransit van)
• 0.8% were on motorcycles or scooters
• 1.3% were on bicycles
• 2.2% were people on foot*

The people on foot had a very hard time crossing Baxter - 77% were the victims of failure to yeild by motorists.

The prevailing traffic flow was vehicles travelling straight along Baxter (78%), with the occassional turn into/out-of the Mid City Mall or Goddard Ave, and the very occassional straight-through pass from Goddard to the Mall (2.4%).

Foot traffic patterns were different from vehicular traffic. Exactly half the crossings crossed Baxter, and the other half crossed perpendicular to that - 13 people each way. Crossing Baxter was tough. Only one walker actually got a motor vehicle to yield to them in the crosswalk. Seven others either were forced to wait for traffic to ignore their right of way, or run in order to safely cross the intersection - sometimes both! Only three crossings of Baxter had neither of these problems. Two had good luck. The third performed a "green crossing" - a crossing outside the crosswalk to obtain the sweet spot in traffic ("green crossing" is the new positive replacement term for "jaywalking", starting now).

There were 15 trips on bicycle. All but one were operating on the roadway. The other was operating on the sidwalk and crosswalk along Baxter Ave, and this operator was clearly too old to do so legally. Still, this was an impressively high percentage of roadway ridership. 73% of cyclists were following Baxter. None were following Goddard Ave. The rest were making turns of some sort.

Analyzing this intersection for danger, I'm worried about that Baxter crosswalk. When outbound car traffic stopped to turn left into Mid City Mall, there was still enough daylight in the wide, ~20' lane for other cars to pass that car on the right. 14 cars passed on the right in this way - 2.2% of the vehicular volume coming from that direction. The left turners did not always use their turn signals, but other traffic correctly divined their intentions. However, there is another reason motorists could be stopping here: they're yielding to a pedestrian in the marked crosswalk. In that case, the stopped vehicle might obscure the crossing pedestrian, and the overtaking passing-on-right traffic can smash into them without enough time to stop. To me this crash seems like a matter of time.

* = Note that my encoding technique covered sidewalk users who crossed the intersection multiple ways, in an 'L' pattern, as two seperate crossings, so this somewhat overcounts the number of crossings. That's because they had to check multiple lanes of traffic to proceed. So the total tally of pedestrians crossings might be a little high, if you prefer to count these together. However, I don't recall any 'L' shaped crossings, so the total number is probably small.