I've heard about fractional reserve features on EOSIO software in which unused CPU and Bandwidth are free to utilise unless the network reaches actual demand.
How does the network determine how much is available? Say if 70% of token supply is staked, does that mean the remaining 30% is available to be used?
Is this enabled by default for every transaction? As in each transaction at the start of a 24 hour period was using the fractional reserve pool, without users actually knowing the TX's they were performing we're going free, until the reserve is completely utilised and then continuing users are then billed for their CPU and bandwidth?