OP, but what mechanism would you propose to protect the network from the entities that would flood the mempool to increase the fees, and therefore, according to your proposal to maintain 1%, would increase the block size so much that many of the nodes that do not have the hardware and the bandwidth will not have the ability to keep up?
I fail to see what would be the incentive to launch this kind of attack as the attacker can't generate more income by increasing the block size limit.
Total fee that a transaction pays is dependent on the size, so it doesn't make sense to peg it to the avg TX fees. You would probably be looking to peg it to the fee rates, but that is assuming that it doesn't get manipulated by miners intentionally including smaller fee TXes or vice versa.
I don't think a dynamic block size really matters. It would be far more prudent to increase the current block size than have to add additional mechanisms to regulate the block size. We're talking about block limits, so if there isn't enough transactions to include, then there is no need for the blocks to at their limits.
You need 51% of the network's hashrate to manipulate the block chain. A miner with 51% of the hashrate could already manipulate the transaction fees.