How do you see the original justification of the n^2 vote power curve as disinsentivising self voting, and sock puppet voting? This still seems relevant.

I'll reproduce a section here from a great post by @bitcoindoom which supports the n^2 curve, reshared on @dantheman 's blog so I presume it has his endorsement also, for what's it's worth:

Under a linear reward model the logical choice is to vote on your own posts as often as you can. This is the other half of sybil only with "fake posts". To prevent this Steem introduced voting power which rate-limits how often someone can vote.

After eliminating fake posts we are still left with the fact that the logical choice (to maximize personal return) is to vote only on your own content as often as the rate limiting will allow you. This is almost identical to paying people interest which in turn would not redistribute funds and therefore would be pointless. Some people may choose to vote for others, but in this case the incentives are identical to a tipping model and subject to similar mental costs as micro payment systems.

So we need to discourage people from voting on their "own" content and encourage them to vote on other people's content. Unfortunately, we have no way of knowing whether two accounts really belong to the same person. The more STEEM POWER an account has the more unique it is and the less likely it is to be fake. Likewise, the more collective votes a post receives the less likely it is to be fake.

Steem introduced the super-linear (n^2) reward system as a natural way of ensuring rewards only go to posts that have strong consensus. Some people may vote for themselves, but the rate of return will be low enough due to being low on the n^2 curve that it is an acceptable cost of doing business. To put it in terms of "micro payments", imagine that you were given the choice to pay yourself $0.01 or to tip someone $1.00. For most people it isn't worth it to work for $0.01, but it would be worth it to them to see someone else receive $1.00. The value of giving $1.00 is much greater than the opportunity cost of $0.01. This is especially true for smaller stakeholders whose vote is unable to climb above the $0.02 threshold for payout. In these cases the cost is $0.00.

The post then goes on to justify down votes to deal with supposed "mining-pools" of voters, i.e. whale collusion.

But anyway, on the curves, do you this this argument is no longer relevant or never was?