If it is just training to become better at the test to measure ADHD (I don't know enough about the game to be sure) than it would be a classic application of that "law".
The gold standard ADHD diagnosis would involve examining the child (or adult) in multiple settings with multiple people's evidence, and then comparing symptoms against those for anxiety, depression, and other MH diagnoses. If they have enough symptoms from the official list, and they've been carefully examined, they may qualify. A simple game/test will never be able to replace that process.
I was originally diagnosed in college. And have managed it with a combo of effort and medication. Life has been good.
Now I have two family member adolescents. One was diagnosed, and I knew he absolutely did not have it. It took two years before the psychiatrist admitted it was anxiety. Meanwhile the younger one definitely does, but personally I feel he is too young for medication or even be diagnosed. He will eventually be properly diagnosed and be prescribed.
I know there is a lot of misdiagnosis, but sometimes it's incredibly obvious even on a first interaction. I am not a psychiatrist but sometimes you just recognize your people.
So like, your gold standard diagnosis ends up depending a lot on the organizational and attention demands the patient is required to meet. A business analyst who is required to sit at a desk for eight hours a day and pay attention to boring numbers on a business spreadsheet is much more likely to get diagnosed with ADHD, while a barber that hangs out and does short haircuts while interacting with clients is much less likely to merit such diagnosis.
This isn't being advertised as more substantive than dual n-back, just more gamified and playable. Given that n-back doesn't seem to work even if you're absolutely diligent about doing it, my expectation for "less rigorous but more playable" is that it will have no value whatsoever.
About the game:
"To meet the objective, users have to identify different combinations of number strings in missions littered with distraction."
About the test (used to prove the game does improve cognitive abilities):
"Those tested were asked to detect digit sequences (such as 2-4 or 6-8), and then hit a button once detected as quickly as they could — and multiple sequences could appear at the same time"
This feels like a paid article to promote the Peak app. No science here unfortunately. I don't understand how public funds pay for studies like this.