Goddamnit no. All of the critics are wrong. PoW coins are not created out of thin air. It costs thousands of dollars in computing power to mine a single block, and that doesn't even include the computing costs of all the miners that didn't win the block. That is real, intrinsic value.
Sort of. First off, the cost to mine a block does not exclude the miners who didn't get a block. They'll get another block, covering their costs.
Second, there is no direct link between the cost to mine a block and the value of the block, except insofar as the market creates one. That's essentially Satoshi's model. Instead of trying to peg the value of the coin to some intrinsic value, you set the value of a block at 50 otherwise-worthless credits, and let the market adjust.
But in theory, the market can adjust to just about anything, as other coins have shown. As long as the coin doesn't fail altogether, then the market has adjusted to its reward schedule, and non-failed coins with all sorts of reward structures exist. So you can't really argue the reward structure must be one way or another on this basis.
The argument for non-trivial perpetual rewards is that coins without them would actually fail. Miners wouldn't mine, or would only do so under conditions that are otherwise viewed as dysfunctional, such as monopolization and transaction hoarding. That's somewhat speculative, but I do think it is pretty clearly possible to set up a coin that fails. If it is possible to do by design it is also possible to do so by mistake, and perhaps easier.
I agree with this for the most part. I just don't like the comparison to fiat money printing (out of thin air), as PoW coins have an inherent effort/reward relationship that is consistent with historical currencies and establishes trust value if nothing else.