> Right now, most of the mining is financed via inflation. But as this comes to an end, eventually, the cost of mining will be borne by anyone making transactions on the network through tx fees.
> Somebody has to pay the electricity bill on all these ASICs.
This is just wrong.
If the award available to miners decreases due to lack of inflation they will use less power-consuming hardware.
If that still results in loss due to electricity cost then miners will leave the market until the amount of miners in the market is equal to the amount of tx-fees available.
In other words: This is a market with supply and demand. If one decreases the other also goes down until they're balanced. It won't just make the market disappear!
So even if almost nobody was willing to pay any tx-fees and there was no inflation then Bitcoin would still be running. It may just move back to running in the background on consumer hardware which was bought for other purposes instead of having giant mining data centers.
This also means that mining isn't going to infinitely waste energy:
The demand for energy cannot go higher than the offer of tx fees / block reward.
So there is a finite upper boundary to Bitcoin's power consumption. I think a finite upper boundary is enough to justify its existence, the precise value of such a constant is arbitrary so you might as well not waste your time in arguing if it is too high or too low and instead be happy that there IS a boundary :)