It is an estimation of the ICC. For example, if we use systematic sampling and ICC = 0 (as estimation), them the variance of sample mean with systematic sampling of single start is very similar or equal to the variance of sample mean with simple random sampling with replacement of the same sample size n.
I believe there will always be to some extent a degree of uncertainty concerning the magnitude of the ICC unless previous studies are available on the same topic (outcome) in the same or similar population. Here "same" and "similar" population is to be thought of, with the outcome in mind. Take a look on this paper, it might help depending on what your research outcome is.