The point of the article is that bikes, by virtue of their limited speed, don't suffer the feedback loop of less density that cars do. The distance you can travel by bike in half an hour is not influenced that much by infrastructure.
In that portion my speed was probably halved, congestion definitely does have an effect at high enough volumes. If more roads in the area had bike lanes this wouldn't have happened and my commute would have been faster (and more pleasant, riding in bicycle commuter traffic is somewhat stressful since everyone wants to go at different speeds). On the other hand if the same number of people were in cars in the same amount of space traffic would definitely have been much slower.
On the flip side, this isn't a city where a large percentage of people commuted by bike. If you turned all the people commuting by car and public transits into cyclists it wouldn't surprise me to find out there were still areas with congestion.
Good bike infrastructure can make or break a commute in speed and safety
The other issue is that infrastructure doesn't take downgrades into account. So, while it can be safe to get up to speeds exceeding 20 mph (30 km/h) while in the center regular travel lane with a significant space cushion due to it's width, such speeds are hazardous with infrastructure. This means that with rolling terrain, the average speed will decrease significantly because one cannot safely travel at higher speeds on descents to make up for the time lost on ascents.