You can dismiss a study with 22 participants as unscientific, if they could have reasonably done a study with many more participants, or if someone else has already done so.
In this case, it's an fMRI study. fMRI access is expensive, and analysing results is time consuming. It's reasonable to start small. If the smaller study comes up with anything interesting, then you can do a larger study later to confirm, and the experience you gained in the smaller study will help you design the larger one.