I already changed to "totally unrealistic" before you commented, but yes, I believe that 30% when the prevalence is at 5% for high estimates is so far beyond reality that it can be laughed at.
Small groups can differ significantly from the population average more easily than large groups.
So they decide they should go on to work with software since it keeps distracting them from everything else.
Ironic that they end up building even more psychologically exploitative software that hijacks our attention