A new kind of crypto-economics

Proof-of-Stake consensus algorithms have been proposed as a replacement for Proof-of-Work. This prompts an interesting question: How many transaction validators should there be in a PoS blockchain?

The tradeoff is between trustworthiness and performance. Consider a PoS consensus algorithm offering Byzantine Fault Tolerance (BFT) properties. If the number of validators is high, the blockchain would be more trustworthy but would offer lower throughputs. If the number of validators is low, the blockchain would be less trustworthy but would offer higher throughputs.

I propose a new kind of crypto-economics wherein there are two native asset types in the blockchain; one of them is liquid and the other illiquid. Market dynamics determines the number of units of each type of asset and consequently the level of decentralization in the blockchain.

I would highly appreciate comments from Etherians. Thanks!