If your primary goal is maximizing bandwidth and minimizing latency, you’re likely building a networked database, not a blockchain.
That’s because bandwidth and latency aren’t core blockchain foundational facets. While speed and scalability do matter, they are secondary to deeper mechanism related principles that define blockchains and their applications focus.
Trying to use blockchains to replace databases means you’ll end up bloating throughput and manufacturing a bandwidth/latency problem that didn’t need to exist, chasing performance instead of prioritizing credibly neutral decentralization. And you will face brutal competition from the world’s best companies in this space: Oracle, Amazon, Red Hat and Microsoft, to name a few.
Distributed databases have long delivered faster, cheaper transaction with the help of central coordination. That’s not what blockchains are for.