The security of this whole arrangement has so far been working good as well.

In order for someone to try and perform a 51% attack, they’ll need to either compromise a large swathe of existing miners (e.g if the government seized control) or create/acquire hardware totaling more than 100% of the existing network today plus growth while you attempt to build more than 100% and then maintain growth over the rest of the network.

As the network grows that becomes exceedingly more difficult to perform.

I have really high hopes for something like proof of work stake, but it’s not without it’s own problems either, and with Ethereum, it’s the first massive scale test, so it’s not as battle tested as proof of work yet, although it’s been used in smaller projects so there has been some testing. With more money on the line though, comes more will to try and break it, or use an exploit you may have held back beforehand.

One interesting difference with POW/POS is that if a miner/entity does somehow perform an attack, they keep the hardware and can continue to try. With POS, they should get slashed in which case the money is gone. But with POW you have the barrier of actually acquiring the correct amount of hardware, meanwhile in POS, you just need the money so there’s no manufacturing/lead time and will be easier to achieve by state actors.