I've copy-pasted some random parts of this awesome article. Everything is taken out of context and probably don't make sense to you if you don't read the original. It's just me writing down notes for myself.
All proposed PoW-alternatives could be relabeled “obscured proof of work”.
Ultimately, each individual “miner” (or whatever) will always:  calculate their expected benefit,  subtract their costs, and  “work” to fight over the difference.
We will re-use the PoW as a heater… First, reusing work is always a smart idea; second, humans do need heat. Third, such an arrangement is highly conducive to Bitcoin’s goal (of achieving a Bittorrent-like resistance to coercive manipulation). However, what is the effect on the MC and MR? Instead of an expected ($100 BTC) per x hashes, we’d move to ($100 BTC) + ($5 worth of heat) per x hashes. If $100 were previously being spent (at the hardware-efficient frontier), spending would be drawn upward toward $105 (eventually Bitcoin’s difficulty would adjust). Inefficient miners (those who didn’t use their waste heat) would be put out of business, but the total spending would still equal $100 (producing $95 BTC + $5 Heat).
The second claim, that the locking-up of funds produces socially-beneficial price-deflation, is wishful thinking. It is not true that “less money supply is available for transactional purposes”; such a statement would describe destroyed security deposits. The only reason the validators freeze their money is to take it out later and buy something.
If 99.9% of the money supply were bonded, prices should fall by a factor of 1/1000 (as they unambiguously would if 99.9% of the money supply were irrevocably destroyed). However, anyone with cash (“liquidity”) can now go on an amazing buying spree.
Plutocracy implies same-wealthy-idiots-on-every-decision and inevitable decadence [unlike “rule of the experts” capitalism], that DPoS strongly concentrates wealth [and, therefore, protocol-control], that voters [even at the more-efficient corporate governance level] never seem to know or care about the issues they’re voting on, and that any conspiracy of 67 delegates can profitably double-spend without getting caught.
Votes are bought. Not directly (this would challenge the voter’s treasured self-delusions, and be a focal point for coordinated outrage), but indirectly through tax breaks, welfare, military action, subsidies, research grants, licenses, etc. Loudly, it is claimed that these all serve civic purposes (the reality is “some”), but the quieter subtext is: “a vote for Candidate X is good for your bottom line”. On top of that, there’s ad-spending and professional campaign management. Because election-competition is zero-sum, these activities cancel each other out and are net “wasted work” for society.
If 3 people vote, a single vote can swing a possible tie, but if 10,000,001 people vote, a tie probably won’t even occur (making each individual vote completely irrelevant). The voting process can be defeated entirely, by taking a small “critical mass of uninformed voters” and, to it, adding some self-fulfilling expectations of hopelessness.
If informed voting were cheap, these self-fulfilling expectations would be expensive to generate. Unfortunately, voting is not cheap, voting costs energy that most people won’t spend. This small detail means that any politician who can quietly use public resources to finance their bribes will always have an overwhelming advantage.
DPoS, like all voting, is not capitalistic. Capitalism is not about “one dollar, one vote”, it is instead “one dollar risked, one vote”. Capitalism requires the potential for gain as well as the risk of loss.
If learning who to vote for takes “work”, then, even absent bribes, voters will always be susceptible to “work”-based influence.
Attempts to create a barrier between spending and result always fail, any leak in the work-barrier leads to total failure (recall that PoW-hashes are individually unlikely to win a block).
Voting works just fine (even very well) when the cost to casting an informed vote is zero, and the votes are aggregated appropriately (mean for a yes/no decision, median for values over a range). If votes are very easy to cast (from a smartphone or something), and the votes are only on preferences/values/priorities (which come from your own head, for free), voting will produce some reliable signal about which course of action would hurt the fewest people’s feelings. This signal could be a powerful, non-violent, society-organizing tool. However, in our case (of validating a blockchain’s consensus, amid attackers who will specialize in confusing voters), an informed vote is always going to be expensive.
In order to ever distribute money in a way that does not produce endless zero-sum competition, the distribution has to be wholly indifferent to human effort.
The 100% premine [of Ripple] takes anyone who wants to build something new, and erases their entire benefit for doing so in a way compatible with the existing system. It takes the creative innovators and kicks them out, forcing them to compete for winner-take-all network effects (…you might as well grab a net and head to Southern Australia if you want a Black Swan this badly). This competition for dominant chain is the Proof of Work, a version of stake grinding with even less focus and an ochlocratic / marketing-based security.
Blockchain security is not the main function of Bitcoin’s PoW. Instead, PoW serves to delay the release of the coins such they they are still cheaply available when potential-network-joiners first discover the project (what some complain as today’s problematic “dumping” by miners). Everyone wishes they found Bitcoin earlier, but earliest-take-all implies the network can’t grow.