Note that the way the article represents the
original study is somewhat misleading.
It's true that the e-bike group reported higher exercise numbers, but that's only when factoring overall energy expenditures from every type of activity (so also regular exercise, energy spent during work, normal cycling, etc).
When looking at just the energy expenditure for traveling with e-bikes versus cycling, cycling slightly wins out (and note that this is just for commutes, recreational cycling is not included). The e-bike group just makes up for this difference in other ways, including by unassisted cycling and larger energy expenditure during work, which leads to higher overall results.
The actual details don't really matter for the big picture, because the results are so close aren't significant from a statistical point of view. So both groups get basically the same amount of exercise. But this type of shoddy scientific reporting is a pet-peeve of mine, so I'll call it out when I see it.