perf: optimize via multicall of storage slot

[rkdud007]

currently provider implemented get_proof method to be accept single or none slot. ideally we should be able to optimize via aggregate multiple slots of same block number + address if it exist.