Okay, problems.
Firstly, they should definitely have more twins than this, right? A sample size needs to be 25, minimum to have any statistical value.
Second, they should switch the brothers. One is clearly into it, and the other isn't, then this is totally going to skew their results with both boys performing worse in life than they should have.
Third, because the brothers have such different personalities, this entire premise fails because the control group is unreliable. You can't determine what the other brother would have become like because they would have inherently grown up to be fairly different.
Lastly, it's too extreme to have one milked like crazy and the other denied completely. Health is often about balance, and making both experiments set to the extreme is only gonna ruin development for both. The proper comparison would be "moderate/mild vs high/excessive".