Forcing them to split the witness elements across many transactions increases the cost of spamming the network. That would mean fewer people do it. It also means more money for miners.
The biggest win would be that it would effectively kill the image/audio/video inscriptions. Only the text based inscriptions would be practical after being split.
But ultimately the point is to send a message to them that non monetary transactions aren't welcome. But instead Core chose to open the doors for them.
If money for miners is a consideration than encouraging spam should be the policy. You can’t have that argument and be anti spam. Splitting the transactions doesn’t stop images and video. Data is data. If a video has to be compiled from a million transactions it is trivial to design a system that does that.
From my perspective I much rather have data in op return that can be ignored than to have millions of transactions that can’t. I do not believe sending a signal that Bitcoin doesn’t like spam is going to change anything. I don’t think changing this stuff actually matters all that much. Feels like virtue signaling at best.
Imagine what the transaction cost would be on sending a video in a single transaction instead of 1000 transactions. That sounds like a deterrent to me.
You could use the entire block in. Single transaction and you would have a video of less than 4mb at a cost of over 3k usd. To do the same thing in 10000 transactions across several blocks the cost would be 2x to 2.5x more expensive and. cause 30% more chain bloat. In the end it’s a short low quality video that’s impractical beyond novelty. Having it broken into small pieces doesn’t change anything beyond the cost and bloat.
I guess you have to ask yourself if 30% more chain bloat is a reasonable trade off for your virtue signal?
