I like the concept of centralization even less than you in the way it's being presented with banksters calling the shots but we can have
a decentralized system of coordinators without having to bend over to know one. See Ripple and how trust is established
Even a decentralized system is only as strong as its weakest link if it has a single point of failure. See Ripple and how tokens are issued.
The consensus protocol itself is quite interesting and I'm sure that LN has borrowed some pages from it (assuming there is no prior work on which both protocols are based upon) but I still find it hard to trust a payment network that relies on a central corporate entity for both development and token issuance. I'm sure that Ripple has merits of its own but it's not quite what I expect from a cryptocurrency.
What Bitcoin solved by applying PoW to the double-spend problem is something wholly different. It seems wasteful, it looks weird, but it's the only meaningful and secure solution so far that allows transactions without relying on a central entity.
Wow, hold on a second because we never needed PoW in the past and tight loops are to be avoided and spamming is avoided
by transactions fees anyway and what about the waste of energy so it now costs 90KWH of electric to process 250 bytes of data.
Reeving engine with car out of gear is crazy
PoW is not about preventing spam transactions, it's about preventing double-spend attacks. Before PoW came along you couldn't rely on keeping a valid, reliable ledger state without a central entity acting as arbiter. Bitcoin's master stroke is making the network not only fault-tolerant against attackers, but actually using what would usually be an attack -- ie. brute force -- to protect the network from the very same class of attacks.
Like everything this approach comes with a trade-off, which in this case is secure decentralization -- on every aspect -- at the cost of performance. At least to me the electricity cost involved seems like a step upward from the waste of human resources that is banking, but that's up for debate of course. One could argue whether the problem of central entities controlling finance is actually a problem worth solving, but assuming it is, PoW is unfortunately our best bet so far. There are other PoX approaches of course, but their viability is a different discussion.
One can argue about whether decentralized currencies are worth it, but arguing that centralization is what would help cryptocurrencies scale is approaching the problem from the wrong angle.
See above comment but let me add to this because virgin nodes starting up uses DNS to get a list of well known nodes anyway
but you will also note that DNS serves cannot steel coins (Maybe they could thinking about it again) so the white-paper broke
it's number one rule anyway and is government blocked domain names on one server then others would take over like OpenDNS
I'm not sure I quite follow you here. You don't need to rely on DNS to connect to other nodes. You can't use DNS spoofing to fake transactions. Worst case you could probably prevent specific nodes from sending or receiving transactions, but that would require a lot of targeted resources without much impact on Bitcoin as a whole.
Assuming you haven't already, you should really look into how LN works. You might be surprised that it's not quite the plaster you expect it to be.
Here's a good starting point:
https://bitcoinmagazine.com/articles/understanding-the-lightning-network-part-building-a-bidirectional-payment-channel-1464710791/