It was already known, as pointed out in the actual paper, that Amyloid-beta (AB) and Tau, the proteins that have long been implicated in Alzheimer's disease (AD), have prion-like properties which means they exist in an abnormal 3D configuration that causes other AB/tau proteins they interact with to adopt this abnormal configuration.
But they can also polymerize into large neurofibrillary tangles (NFTs). Patients who died late, e.g. at age 80 with AD had a lot of these NFTs, so it seemed reasonable by many to assume that these NFTs are the major cause of brain deterioration. But there were a number of alternative hypotheses of what exactly AB and tau were doing in the disease.
This study suggests that AB and tau cause disease primarily through their prion-like activity and not due to their ability to accumulate into large trash-balls (NFTs). They noted that patients who died young of AD had very few NFTs but have a lot of the AB/tau prion-like forms, whereas patients who survived longer had a lot of NFTs but not a lot of prion activity. This lends evidence to the idea that the primary pathogenicity of AB/tau is through their prion activity and not the fact that they accumulate into large protein blobs.
This is unsurprising, given the last two decades of drug trials targeting these molecules have all uniformly failed.
The alternative hypothesis is that these plaques are protective against neuronal infections that provoke alzheimers. That hypothesis has not garnered wide support yet, though it is growing, and is at least consistent with the inverse relationship observed.