This appears counterintuitive; people typically predict intense states to last longer. The hypothesized for this disconnect is that, intense states trigger psychological defense processes that reduce the distress, while less intense states do not trigger the same psychological defense processes and, therefore, less effective attenuation of the stress occurs.
The article keeps falling back to the bike example, which totally ignores the time required to equip the bike and lock it up at the destination.
Their other example of a bad romantic relationship is also hung up on switching costs.
I prefer the version “broken gets fixed, shoddy gets ignored”.
Some living situations have time-consuming storage places of bikes, like I can imagine some apartment building with cellars for bikes or shit like that.
But locking your bike at the target place? Takes seconds. Also, I’m relatively sure for most people, retrieving the bike also takes at most a minute.
Now look at the graph. Shift the bike 1.5 minutes upwards, you still get almost exactly the same effect, the bike is still faster for a lot of the 1 mile of walking.
So sure, it’s important to take into account the switching costs, but this concept still holds true most of the time.