Bitcoins are the consistently valuable and liquid coin, therefore it can make sense to serve bitcoin mining js.
As it is javascript is the number one reason for shit battery life, I don't want to make that even worse than the days of flash which its rapidly already become today.
Faustian bargains like that aren't a great definition of progress.
Fundamentally the issue here is energy consumption while on battery. Without a way to say: a site can use N% of battery and then hard cap the usage down to a trickle and alerting the site that the user has limited energy resources available, I think this is a non starter.
Similar to when you are on LTE you don't want to use up limited bandwidth for say 4k video and want to say: you know what, I'll deal with this later and not go over my monthly allocation, the same applies here.
Is that view that unreasonable?
Why? Because each desktop CPU would only yield 20-30 megahashes per second, less if you don't want to slow the user's computer to a crawl. A server would need to manage millions of clients just to match a single ASIC miner which has efficiency rated in terahashes. And that amount of hash power would only return about 30 dollars in bitcoin on a good day.
This is why all javascript miners target Monero and the latter was often criticised for having a CPU-friendly PoW because this is just asking for botnets to form.