How to prove improvement in transaction throughput

[wowfacftory]

I am working on a research in bitcoin blockchain where I am making some hypothetical changes in policies of bitcoin blockchain. Due to this, it helps to increase the transaction throughput of bitcoin blockchain (or at least a thing I want to claim through my research). Or one can think of this research as an identical PoW blockchain to bitcoin, with some changes in its policies and structure. I am already working on a simulator called btcd - https://github.com/btcsuite/btcd. It is a full node implementation of bitcoin blockchain or what I understand of it. I was wondering how can I prove using btcd that transaction throughput have been increased through new policies. Other possible simulator names with suitable abstraction level will also be helpful. If we think of a scenario where it is absolutely necessary for miners to fill the blocks at its fullest (Apparently,  It will improve transaction throughput.), can this improvement be proved using any mathematical aspect?

[joniboini]

You can probably prove it by:
- Comparing how many transactions are included in each block.
- Comparing how long it takes to clear a clogged mempool.
- Comparing required fees to confirm a transaction (by assuming that if you require a very high fee then it means throughput is low).