Bitcoin has merits. So proof of work, being a part that is currently necessary for it to work has merits.
But the argument that proof-of-work has the merit of incentivizing cheaper energy sources does not stand up.
1. All uses of energy already incentivize the search for cheaper energy. This isn't a novel incentive that proof-of-work provides.
2. But proof-of-work does have a relatively novel disadvantage. It won't just incentivize greater energy uses as prices come down, as in normal supply-demand curves, but must keep up the original and even grow the amount spent on energy.
This would not be a problem in a market without negative externalities, but energy is famously a huge industry that will be struggling with negative externalities for quite some time and at great cost.
All uses of energy incentivize the search for cheaper energy.
Not all energy uses are optimized to increase usage over time, to cancel efficiencies. Proof of work, whatever it's merits, is anti-efficiency with respect to itself.
This is a significant difference.