pull down to refresh

I don't think it makes a significant difference to forward the transactions with the blocks or when they're relaying unconfirmed. The node might save a few inv messages and forwarding a tx once or twice. If that bandwidth savings is of concern, blocksonly would be more effective at reducing bandwidth use.
I don't see any non-marginal benefits to be achieved by filtering unconfirmed inscription txs beyond virtue signaling, and I don't think working on enabling virtue signaling is a good use of my time.
No one's making you do anything. You are though, volunteering your time to voice your opposition to what you view as virtue signalling. While parroting the same tired false goal post I make a point of calling out in my article. baffling.
I view this anti-spam activity as reducing moral hazard and overall risk, while preparing for the worst case scenario.
RAM, CPU, Bandwidth and energy costs are all of concern here.
If normal people can't run a node because of this activity in 14 years, we have a problem.