over time the value of the electricity consumed by the miners will approach the value of the total block reward of the blocks.
This is a partially concerning part, but it may also be comforting. I'd need to do the math on it first to decide.
After halving, Block reward may as well be at zero. So let's ignore it for the moment. Over the long term, electricity consumption will be dictated by fees paid by users. The total fee (# of txs * average fee) will need to at minimum = the cost of electricity.
A couple things can happen with this:
1. Bitcoins price increases dramatically and covers the ever increasing electricity demands
2. We reach an efficiency bottleneck with ASICs and electricity consumption requirements level off (seems unlikely and also you could just manufacture more asics - so I think this ones out)
3. Average fee price increases to meet the increased demand of electricity
4. # of txs increases to increase the total fees paid
5. Bitcoin doesn't increase in value, fees and txs don't increase, miners hit a bottleneck due to the cost of electricity and profitability. Bitcoin becomes less secure.
You are missing an important parameter: total hashpower (or hash difficulty beause they are linked) that can flucuate depending on how profitable mining is
To sum up all parameters for a simple model you have: average fee, number of transactions per block (can be taken constant), block reward, total hash power, electricity cost
The difficulty is set so the total haspower mines a block every 10minutes
Do the math doc
What you will be missing then is how the total hashpower increases or decreases depending on the average profit (some miners will quit if their profits are to low and more miners come if profits are high). You can make your own model and find a nice equilibrium.
Too many miners quitting due to loss of profitability = bitcoin security degraded = attacks more likely and bitcoin weaker.
Bad scenario - miner profits drop significantly, miners cannot continue mining, miners sell mining equipment at a discount en masse, a malicious party/government buy up surplus mining equipment, and mine as an attack, not for profit.
That's the danger of the above.
You're right that mining finds an equilibrium, but if that equilibrium is, for example, 50% of previous hash power, that's a HUGE problem.