Regarding under- and over-eating: it seems you think I am a simple mathematical average that doesn't factor that in. I stand by my observation, as and observation, not fact.
Lisdexamfetamine is not amphetamine—not chemically, not in terms of its half-life, not in subjective experience, and not in any study that tracks behavioral or long-term effects. At best, it's a prodrug of an enantiomer of amphetamine.
You are also mistaken about the study. Reading even the abstract would clear that up for you.
Let me clean up the unnecessary, convoluted language before I answer:
Q: Does it stop being called "abuse" once a doctor prescribes it?
A: No. The prescription stopped hospitalizations due to amphetamine overdose.
Q: Is it simply safer to use drugs with a doctor's help?
A: That was not answerable based on the study's design. It is also not a useful question to ask in this context, since it's comparing apples and oranges. Some of the worst cases of drug abuse are created and maintained by doctors. However, taking drugs collaboratively with a doctor is probably safer on average than getting them from random webpages.
Q: Is this specific amphetamine safer than others?
A: Yes, as is the case with any substance we ingest. You can quibble over the details, but beer is safer than hard liquor. Likewise, different medications in the same category or receptor affinity group have different LD_{50} doses (the ratio of the clinically effective threshold to the threshold where 50% of subjects would die).