I'm not into Ethereum, but "more precise estimation of gas costs" sounds like a letdown compared to what you would want to build a really reliable smart contract. Can't a contract just die if it can't pay its gas bill?
With all the investment in blockchain (as you yourself mentioned in another comment), and so much money riding on the correctness of contracts, it seems like the incentives should line up to develop proper, usable formal methods (starting with backing down from Turing-completeness for the most part, most likely). That would be a pretty nice side effect for us skeptics. That's honestly what I was hoping for, with OP.
Contract’s don’t (more accurately almost never) pay for gas. Gas estimation is done when sending by the end user.
Some parts of gas estimation can be done statically, but some of the cost of certain operations are dependent on the data being loaded from the state. I’d like a tool to set an upper bound on gas usage (current approach is to execute the tx on pending state and add 10-30%), but solidity does not provide one.
With all the investment in blockchain (as you yourself mentioned in another comment), and so much money riding on the correctness of contracts, it seems like the incentives should line up to develop proper, usable formal methods (starting with backing down from Turing-completeness for the most part, most likely). That would be a pretty nice side effect for us skeptics. That's honestly what I was hoping for, with OP.