- The cost of running an ETh2 node is independent of the amount of Eth staked (electricity, networking, and hardware costs don't care if you have 1 ETH or 100 staked)
- The amount a node operator makes is a % of the staked amount.
With those points in mind, it would seem to me that node operators will be more profitable if there are higher staking requirements.
For example, if it costs me 0.2ETH a year (at current prices) to run a node, and the staking requirement is 4 ETH,
I might only make break even (depending on the actual % return rate and how much Eth is staked in all of Eth2).
But if you doubled the stake requirement to 8ETH you've now doubled the return that node operators receive while not changing their costs.
It would be interesting to talk about some protocol consensus method for adjusting the staking amount. It seems like most protocol discussions are about having a certain amount of ETH staked, but I think the number of nodes should also factor in as well.
Raising the stake amount would lower the number of nodes but it might actually allow nodes to be run in areas of higher electricity/network/hardware costs and be more decentralized.
People who want to stake but do not have enough ETH to run their own nodes could still pool their stake with other node operators so they would not be excluded by higher staking requirements.