Broadcast, print, and electronic media have carried quite a few reports recently regarding how long coronavirus will survive on various surfaces (paper mail, food packaging, doorknobs etc). I have no special expertise in virology or viral biology, but my gut impression is that coronavirus primarily spreads like other respiratory viruses, via droplet nuclei and aerosolized particles, and that transmission by contact with fomites (physical objects) is a relatively unimportant factor in its epidemiology.
I no longer have access to the primary data on most medical articles, relying like everyone else on second- or third-hand interpretations of what a study discovered. But yesterday I decided to chase one article down. A report in LiveScience was titled "Coronavirus can survive on skin for 9 hours." The study was published in Clinical Infectious Diseases, which I understand to be a reputable journal, and the original data is available in pdf form.
The image embedded at the top is Figure 1 from the article - a composite illustration of the methodology and the results. I don't question the findings per se, but I have some concerns re the extrapolation to real life. Note that the bar graph shows a decline in detectable virus (FFU=focus forming units, equivalent to colony forming units in bacterial studies) from about 10^5 at time zero to 10^2 one hour later. Clearly the virus did survive that interval, but it incurred a 3-log-power decline in that time.
So how does this translate to real-life risk? I don't know what constitutes a normal viral inoculum. How many FFU are there in an aerosolized droplet? On an infected person's hand when they touch a doorknob or someone else's hand? The authors seems to acknowledge this in their discussion: "... since not only the virus stability but also the infectious dose and transmission route may greatly affect the risk of contact-transmission, future research needs to focus on factors other than virus stability."
After three decades spent writing and reviewing research grants and manuscripts for publication, I know that the main conclusion of most research is always that more research needs to be done. Some of it may have been done already; if any readers know of relevant results, I'd be delighted to see the info in the comments section of his post.