Here's the actual article:
https://www.medrxiv.org/content/10.1101/2020.03.09.20033217v...I would not read a lot into the "awaiting peer review" bit. It normally takes at least a few months for an article to be reviewed, revised (if necessary) and re-reviewed. Things are undoubtedly fast-tracked for COVID-related stuff, but you can only do so much to speed it along.
Per their Table 1, the median halflife of an aerosol is about 3 hours, with a 95% credible interval of 1.81 of 5.45 hours, so it's not a a best (worst?) case number at all.
The plastic data in Figure 2 shows that it does decrease a lot by 48 hours (~1000x), but it's still above the limit of detection; they estimate the half-life to be 14.8 hrs. The 3 day claim is true, but is more of a best/worst case scenario