If core decides to merge the OP_RETURN data limit, I won't update to any new version of core with this default implemented.

Just think about what use cases coule and which most probably will emerge.

As an example you can broadcast data through the mempool without paying for the transmitted data by RBF'ing the tx before the tx hit's the first confirmation and replace it as long as it doesn't get that much sats/vb to be confirmed. This lowers the relative fee for any further RBF (with new data included and distributed) DRAMATICALLY and will create incentives for mempool only use cases which just consume node bandwith which isn't even paid for to the miners. Sounds like siphoning the avaiable bandwith of node runners for the distribution of data. I won't give my node ressources in support of creating and maintaining a relative RBF sats/vb market.