How is this an attack to the network? Exploiting the fee market is not quite an attack or is it? I'm not entirely convinced of the benefit of this hashrate gauging that philipma1957 describes, but there're some transaction market variables which are to some extend beyond the control of mining pools, e.g. transaction volume of bitcoiners induces by external factors like Bitcoin market value development.

As far as I remember the last time we didn't have mempool congestion was around midth of April 2023 and before that at the beginning of 2023. Since then mempool never got emptied enough that there wasn't an excess of transactions waiting to be confirmed in new blocks.

When mempool has more unconfirmed transactions than fit in a (few) block(s) there's natural fee rate competition, a natural fee rate market for bitcoiners. To maximize benefit mining pools should choose transactions solely based on the max. fee rate they offer, unless the pools have some special paid services, like paid transaction accelerators, that can compensate choice of lower fee rates.

Anything beyond the beneficial choice of transactions based on their top fee rate is likely intransparent, trust based, manipulated and hardly verifiable for participating miners of pools.

Concentration of hashpower in large pools is a problem of concentration of choice of which transactions go into blocks of those pools. Usually the pool decides about the block template that participating miners work on and miners don't have a word on the block template. Miners only choice is at which pool(s) they mine. Most likely this will be a pool which maximizes their profit. Mining gear and energy have to be paid...