Reconsider rosetta new block timeout length after inactive stake protocol changes, slot time reduction

The PR https://github.com/MinaProtocol/mina/pull/18526 is going to be increasing the timeout in the rosetta connectivity test when waiting for new blocks to 40 minutes, which should make the test fail far less often. This was necessary because of our mainnet fill rate issues. We could potentially adjust this timeout, though: - Mesa will halve the slot time, so this timeout is going to become twice as generous after the hard fork - If we do implement the inactive stake protocol changes, the fill rate on mainnet will improve Either one might warrant a decrease in the timeout.