Solving for state bloat in custom L2 implementations

I’ve been diving deep into some research for a custom Layer 2 scaling solution, and I’m running into some serious walls regarding state bloat and data availability. Most of the standard documentation is great for 'hello world' dApps, but it gets really thin when you’re trying to optimize at the protocol level. I’m trying to find a group that doesn't just write smart contracts but actually understands decentralized systems at a fundamental, first-principles level. Has anyone here collaborated with an external R&D group for protocol-level engineering? I need some heavy-duty engineering eyes on this.