Bitcoin is just fine but talking about ways to improve it by relaxing some requirements in favor of a simpler behavior (such as using some external time references widely available over the internet) i think always helps.
I have to disagree with you, but this part is a matter of opinion.
I think Bitcoin should strive to be as decentralized as possible. It's not perfectly decentralized now, but I see no reason to add additional potential points of failure or points of control. Rather I'd prefer the opposite -- remove what few remaining centralized points are left.
It is obvious that you do not forward anything received before you double check it and compare it in case of block hash with your own best result. So a modest performance node can only flood minor performance nodes.
Sorry if i did not mention that but i was aware of this practice and should not change it.
I suspect that the majority of nodes on the network are not miners, and therefore have no concept of best-block-so-far-in-my-pool, and remain susceptible to a DoS.
What is the max range of possible bit transitions? 128
Is it easy to generate them? not so straight
Is it easy to count them? yes simpler than generating a random one of them or calculating a hash
I think you just described a PoW system... hard to generate and easy to verify. How this add-on PoW is any improvement of the PoW already in use?