There are degrees of uncertainty, and ranges of plausible severity, even when we haven't actually nailed these down quantitatively. We already have enough information about COVID-19 to raise it far above the threat level of a random virus. And given how quickly and stealthily it seems to be capable of spreading, to take no action beyond 'pursuing more information' would be almost the same as assuming the best possible resolution of our uncertainty. That's no more rational than the other extreme of assuming the very worst.