Subnet Optimization ideation. Innovation in the execution and consensus layers within a Subnet VM

I have recently looking at parallel optimization methods from Solana and Monad and other implementations being done. I see that they are all lacking security and composability. As Avlanche is considered "fast enough and horizontally scalable which I see’ would like to be the devils advocate. I see it’s strengths and potential threats

I get how Subnets are supposed to be designed towards application specific usecases but thought I’d throw this idea out there and ask if anyone is iterating on it already…

  • The ultimate Subnet scalable subnet
    A Subnet that can batch transactions in parallel with common dependencies. A account state tree optimized for concurrent processing. A consesnsu mechanism where validators aggree on the contents of the block that will be executed. Instead of giving block interval time to nodes to communicate across the world. Why not give that time towards execution time. And let the nodes decide on queue. Reducing overlaps makes tons of sense.

All food for thought I am a big fan already of the approach AvaLabs and Core team has been taking.