It sounds like a normal selfish-mining attack, that we've known about for a decade, but with a new analysis that suggests it's to every pool's advantage to do it.
Ah, the article you link to in your newsletter is just slop. The attack is described well in the paper:
An attacker is either a solo miner or the manager of a mining pool. The attacker can create numerous Bitcoin accounts with different identities, and control them to join (infiltrate) other open pools. However, we assume that attackers cannot infiltrate those private pools requiring private membership. The attacker’s computational power is limited, yet it can be dynamically divided into two fractions used for innocent mining (i.e., working as an honest solo miner) and infiltration mining (i.e., joining and mining in multiple open pools to gain extra illicit rewards), respectively. Meanwhile, the mining power for running the attack (e.g., infiltration mining or selfish mining) in an open pool should be “loyal”2 [18, 32] and the fraction of it is a secret [18].
tldr the attack involves a large miner selfishly mining on multiple pools.
It sounds like a normal selfish-mining attack, that we've known about for a decade, but with a new analysis that suggests it's to every pool's advantage to do it.