When coughing or shouting, recent systematic reviews have shown respiratory droplets can travel more than a couple metres. In one study, a violent exhalation of air spread some droplets eight metres away (26 feet) in just a few seconds.
A one- or two-metre rule could very well be sufficient in some situations, but scientists in the United Kingdom say we need a more nuanced model.
Right now, they explain, the rules we have don't take into account subtle factors like ventilation, time spent together, indoor or outdoor settings, mask use, or the type of social activity occurring - all of which could impact the spread of the coronavirus.
What's more, distancing rules often don't consider the size of airborne droplets, how much virus the droplets can carry, or how susceptible others are to these viral loads.
Critics of stricter measures say we are being too cautious, and while that's probably true in some situations, in other cases, scientists argue we are likely not being wary enough.
"Instead of single, fixed physical distance rules, we propose graded recommendations that better reflect the multiple factors that combine to determine risk," the authors of the new analysis write.
"This would provide greater protection in the highest risk settings but also greater freedom in lower risk settings, potentially enabling a return towards normality in some aspects of social and economic life."
The review joins several other recent critiques of current social distancing rules. In July, hundreds of scientists co-wrote a comment piece urging the World Health Organisation (WHO) to reconsider its advice to "maintain at least one metre (three feet) distance between yourself and others."
"We are arguing that there is insufficient proof that aerosol/airborne transmission does not occur."
To what extent that occurs is another matter, but there's mounting evidence the coronavirus is airborne, even in tiny droplets, so the new analysis from the UK takes a similarly prudent approach.
Some recent reviews have found the risk of being infected with COVID-19 within a metre is roughly 13 percent, whereas beyond a metre, it's only 3 percent.
Still, the authors of this new analysis say estimates are based on flawed and often outdated science, some of which goes all the way back to the 1930s. All those decades ago, we predicted how far respiratory droplets can fly when a human coughs or sneezes. Yet that simple model doesn't examine viral load, different sizes of droplets that can travel over a range of distances, or the type of virus itself.
Without exhaled airflow, for instance, large droplets appear to travel at max two metres away, while small ones succumb to drag and evaporation much sooner. With exhaled airflow, on the other hand, clouds of small droplets have been shown to travel beyond two metres.
A study at a hospital in Wuhan, China even found traces of coronavirus hanging in the air roughly four metres away from patients.
Some infectious disease specialists aren't too worried by this, as smaller doses of airborne coronavirus may not pose as big a threat of infection.
A systematic review of social distancing measures, commissioned by the WHO, found that a metre or more of separation could decrease transmission risk by roughly 10 percent. Yet scientists in the UK argue those data are largely based on other coronaviruses, and only partially account for environmental conditions.
While it's hard to trace back individual infections to their precise source and the distance the person was at, there's reason to suspect respiratory droplets might play a part in the current spread of the pandemic - at least in some situations.
In meat packing plants, for instance, outbreaks have been particularly bad, and the authors say this is probably compounded by higher levels of worker contagion, poor ventilation, cramped conditions, background noise (hence, shouting), and not enough mask wearing.
The same sort of conditions might be expected in a pub or a live music venue, they add. We've already seen cluster outbreaks in gyms, call centres and churches, where people talk, pant or sing loudly.
At a choir practice in the United States, one symptomatic person was actually found to have infected at least 32 other singers, and possibly 20 more cases yet to be confirmed, even though the choir members were socially distancing.
These documented outbreaks require an explanation, the authors argue, otherwise they'll just keep happening.
Even as restaurants and bars reopen, countries like the UK are still telling people to stay at least a metre apart, and that could end up misleading the public, making people feel safer than they actually are in riskier situations.
"Physical distancing should be seen as only one part of a wider public health approach to containing the COVID-19 pandemic," the new analysis concludes.
"It should be used in combination with other strategies to reduce transmission risk, including hand washing, regular surface cleaning, protective equipment and face coverings where appropriate, strategies of air hygiene, and isolation of affected individuals."
The study was published in BMJ.