Thank you, both are fair points.
On the count: the HP Distribution report only includes accounts that actually hold Hive Power. Out of the 2.6M registered accounts, a large share hold zero HP, either empty or fully powered down, so there is no bracket to place them in. That is why the total here lands around 1.19M.
On the fairness point, that is a sharp distinction and I take it. The stake really is very concentrated, that part the numbers show clearly. But you are right that HP is not like fiat wealth. It carries the inflation and the price risk directly, and the big holders are usually the ones funding curation, the witnesses, and the proposals that keep the ecosystem running. So "concentrated stake" is accurate, but reading it only as control misses that a lot of it is cost and responsibility. Your framing, that the top accounts carry a large share of the bills, is a fair and useful way to read the same numbers.
RE: How Hive Power Is Spread Across the Network: A Look With HiveScan