Revealed preference. If Buium was really considered top 5, someone would have jumped on him.
I think it comes down to variance of perceptions, that is, he might have had a smaller variance of expectations while other players had larger variance. So on average, would have ranked above players who a minority of GMs saw having higher ceilings.
Since it only takes one GM, if a GM sees someone having a higher upside they may take that player first.
Buium was old (Dec 7 B-day) for his draft class, so the perception may have been that he was closer to his ceiling, even though his mean was higher, than other top prospects.
NJ for example, might have seen Silayev with a higher upside (despite higher risk b/c of limited track record, though was in KHL at 17).
Could be "winner's curse" or just a lower level of risk aversion (more willing to gamble on ceiling).
The interesting choice was Catton at #8, mediocre D-1, great D-0 season, Jan B-day, hadn't filled out at 5'11 172. Was this a "mean" pick or an upside pick?