OP, a lot has been written already about this issue in several threads. Most were however more active in 2023 and early 2024 when the Ordinals wave was at is maximum, when I think you weren't still around in this forum. I'll try to briefly summarize my opinion about these topics in simple terms:
1. Is this exploit done for selfish or general benefits?
There are two problems with Ordinals/Runes.
The first one is the presence of transactions with big amounts of data (up to 4 MB in some cases). This is actually the "exploit" some talk about, because a Taproot "feature" was mis-used to circunvent limits for standard data transactions and thus storing lots of data in a transaction became cheaper. However, these transactions have only led to mempool congestion for a brief time in early 2023.
The second problem has much more incidence in the congestion: the appearance of tokens like BRC-20 and, more recently, Runes. These transactions are small, but their purpose isn't to transfer Bitcoins so they can be seen as "spam" by those wanting to limit the Bitcoin network for "payment" transactions. However, tokens on Bitcoin are nothing new. They exist since 2013 approximately (coloured coins, Omni/mastercoin). And more important: with the Bitcoin protocol it
can't be prevented that such tokens are created (early token transactions looked exactly like normal transactions but had the data embedded in sequence fields and fake public keys).
2. Do you think Ordinals, Runes and BRC-20 will become a massive problem in the present and future?
I believe we will see some waves still, more from Runes than from Ordinals, but they will become much less pronounced. BRC-20 is already almost dying. Runes transactions have seen a second wave now but it's lower than the first wave (near the Halving event in April).
3. should we focus more in resolving them or increasing block size instead?
Neither of both. We should support all investigation which goes into second layers (which is the "transaction batching" aspect some mentioned in your other thread). This is not a fact like the answer to 1, but a personal opinion. I'm generally in the "small blocker" camp.
4. What's the answer to the topic title above?
If you think every data transaction is an "exploit", then probably yes. Data transactions like Ordinals, Runes, and to a lesser extent older protocols like Counterparty, Omni etc. are making up around 50% of the block size currently.
But if you only take into account the Ordinals which used the exploit I mentioned above, then its incidence is low, as I mentioned before.
Only bitcoin developers are the ones that can make this discussion and it is clear that they want the tokens to stay.
Disagree. They are quite neutral about them, but an aggressive development action against OP_RETURN tokens like Runes could cause massive side effects like a bigger UTXO set due to "fake public key tokens" like Stampchain.