As you rightly point out, whether something is considered to cost too much power properly depends on what you compare it with.
Before there was a viable alternative to POW for achieving the goals of a decentralised currency, it was at least debateable about whether the benefits of the decentralised currency were worth the cost of POW compared to the alternatives, but now that it appears POS is a viable alternative, there can be no question - POW costs too much power.
If it turns out PoS is actually a good idea, I think you will have a very good point. Until it's been running and is used right (and doesn't get exploited hard and early) I think it is premature to say this.
Before there was a viable alternative to POW for achieving the goals of a decentralised currency, it was at least debateable about whether the benefits of the decentralised currency were worth the cost of POW compared to the alternatives, but now that it appears POS is a viable alternative, there can be no question - POW costs too much power.