I had only read the first, which I summarized accurately. It shows that high overall turnout doesn't always correlate to dem victory, but explains that this is because turnout only matters in swing states, and dems have gotten very high turnout in states like CA in elections they have lost. That article is totally unconcerning. It's obvious that turnout only matters in the swing states.
Now I just read the second. That one says that a poll of chronic non-voters shows more of them favoring Trump in a few swing states, while in other swing states it's a tie, and nationally it favors the dems but that is just padding the popular vote for dems while doing nothing to help them win the election. This is concerning. However, I read the PDF of the study which showed that they surveyed 12,000 people nationally, meaning that the national number is probably solid, but they polled only a few hundred people in each of those swing states, suggesting a high sampling error for individual states.