One possible optimization to compress Bitcoin transaction size is to avoid repeating redundant data between inputs.
As the agreement changes, new transaction types will allow you to sign the following lists: txid:vout They are paired if they all share the same output script.
Taking an extreme example, we have 1.6 million outputs with the script 1HckjUpRGcrrRAtFaaCAUaGjsPx9oYmLaZ. Using one of these P2PKH outputs costs 40+72vBytes (prevOut + scriptSig) per input. By aggregating scriptSig across inputs, transactions can be included only once per transaction, omitting 72 per input. When combining all these outputs, these optimizations can save approximately 115MvBytes of network data. (At a commission rate of 10 sat/b, this translates to 11.5 BTC, which is approximately 80% of the total value of that output)
What are the pros and cons and hidden problems of this kind of optimization?