Bench-marking research engine speed

Please, I want to make a benchmark to study the speed of the local search engines in blockchain. For some delay-sensitive application, this metric is important to decide about the usefulness of the blockchain or not. More specifically, when a local copy of the blockchain is stored on the local machine, what is the times (in seconds or milliseconds) required to find a particular block. The search can be made by address, block number …
I think there are many things to take in consideration, such as the indexing method, the time of the first indexing, the synchronization, the indexing update and the search. Any advice?