Again, another example of "headline science." In every case of these kinds of studies, it's worth taking a look at the actual studies or at least the abstracts, because the headlines tend to generalize findings that shouldn't be and rarely discuss the limitations to the study that the authors normally address up front.
The NIOSH study was published in 1994 and its sample included only players who had played 5 or more seasons in the NFL and by its own admission included very few men who had reached 50 years of age at the time of the study. If you read
this article in the newspaper of record but didn't click over to the abstract, you wouldn't see such a limitation mentioned.
Findings suggested that this cohort of NFL players had lower mortality rates at the ages they had reached than a general population that were from the same ages. However, it also found that football players, particularly linemen and black players, were at an increased risk of heart disease (number one cause of mortality in the developed world), almost certainly due to body size, and had no difference in incidence of cancer. They were less likely to suffer from accidental death or death by violence than the general population.
They had higher incidence of ALS, too, which could be alarming (if it's not by chance, as the authors suggest) because CTE is hypothesized as one potential cause of ALS. Kevin Turner had ALS and CTE pathology was found in his brain, for instance.
What the study didn't do was compare them to a control group of non-NFL players of similar socioeconomic status. Accidental and violent death are much more common among people of low socioeconomic status, and that's not a category most NFL players who played for five or more seasons fall into after their playing careers.
In other words, up to age 50, it seems NFL players (who played 5 seasons up to 1988) were less likely to die than a similar cohort of the general population. However, they may also be just as likely to die as a similar socioeconomic cohort, and their life expectancy at age 50 (I'm a demographer; we notate this with e50; e0 is life expectancy at birth) may be much lower than the norm. We don't know; there aren't any studies to tell us.
I agree that there's an overdetermination of co-morbidity when it comes to CTE, but the actual effects aren't exactly unknown. The pathology is similar to senile dementia or Alzheimer's, and we know what those do, and as mentioned it's a hypothesized cause of Parkinson's and ALS.
Even people with early onset Alzheimer's or dementia, which this is perhaps most comparable to, can live up to or beyond the expected age at death for someone of their gender and socioeconomic status. It's not just a question of mortality, though that's what grabs the headlines (aided no doubt by the fact that all cases of CTE have been recorded in dead people due to limitations in screening) but a question of morbidity and quality of life.
See Earl Campbell or Muhammad Ali for instance. The former is age 62, the latter died at 74, which is right around life expectancy at birth for a black male. But Ali lived for 30+ years as a walking zombie, diagnosed with Parkinson's (a disease that's hypothesized to be related to CTE, at least as one way it's acquired), unable to care for himself. Campbell may live to be 80, but he's incapable of speaking in complete sentences. It's not just premature death.