Ordinals are forcing us to consider the limits of the Bitcoin mainchain. We had the first 4mb block recently. That is a massive event that it is essential to consider whether we can support. Anyone who cannot currently handle it is making the necessary changes to support blocks of that size. This means that our infrastructure is stronger for when a state actor attack inevitably occurs. We will be able to handle any throughput that they might try a spam attack with. Events like this, while painful in the short term, prepare us.

There is only so much block data, and more of it being used for jpegs, albeit stupid, means that we learn to deal with higher fees and scarcer blockspace. As Bitcoin grows to more common and vital use cases, we will need the infrastructure to scale beyond current limitations, such as drivechains and lightning RGB. Higher fees now means progress in the future.

Halvings will make it harder and harder to get revenue from mining operations. This may or may not become a problem, but it is essential to ensure that mining remains profitable. Because it is customary for fees to be linked to blockspace, more blockspace use means more fee revenue. Thus, if inscriptions need to pay much higher fees, which keeps miner incentives healthy. This effect is increased by the community's dislike of ordinals, as many miners may have to be "bribed", i.e. higher fees, in order to include them.

For a better steelmanning I would look to this article on how ordinals impacts mining, as well as the intro of this episode of the Bitcoin Dad Pod.

Personally, I think Ordinals might have some good use cases eventually, but that right now they are kind of dumb and serve only as a stress test.