At some level of "low" it's next to useless. I hope that when SAGE give a range of values for R they are doing so because the numbers of positive tests are high enough for their figures to be useful.
The ONS figures from random sampling give ranges, and they also say explicitly when the values shouldn't be relied on because they come from too small a number of samples. Recently, they have warned that regional values of R should not be taken seriously but the whole England ones should be valid. Random sampling is from a quite small number of tests so R values from this method will be less accurate.
In any case, if R has become unreliable due to being taken from too small a number of samples, I'd have thought you'd see it fluctuating wildly, wouldn't you? And it doesn't seem to be doing that.
I think it is incorrect to say that currently R is not at all useful.