The time it takes for a network to come to consensus is the time it takes for a block to be propagated and validated across the network. That is a function of block size and network latency/bandwidth.

http://www.tik.ee.ethz.ch/file/49318d3f56c1d525aabf7fda78b23fc0/P2P2013_041.pdf

Block time in general is a decision made as part of the protocol development, weighing the value of granularity of information vs. the fears of centralization where network latency gives an advantage to larger players working together. Ethereum and others could have chosen faster block times from a technical perspective, but weighed them up in this way.

https://blog.ethereum.org/2015/09/14/on-slow-and-fast-block-times/

In my estimation, they are just fears. Every time block time has been reduced or block sizes have been increased they have been found to work well without serious drawbacks.

"Then why are no other blockchains running PoW and capable of processing that many transactions?"

I can't answer that with certainty but my view is that it is a matter of fear. There is a certain culture pervasive in the crypto community which does not seem to be or have been present in the development of Steem. Steem has firmly embraced the idea of nodes being run by institutions from the start, not held back by the expectation that end users must run full nodes to validate everything locally. Most other cryptocurrencies have inherited that from the Bitcoin community. Even Bitcoin Cash has maintained 8MB limits (despite successfully testing 1GB blocks) out of conservatism.

I agree with your points on the financial costs added to the system and the comparative ease of hard forks allowing faster development progress.