So far we can't say anything, because all we're seeing are our own detection limits. This is called an "observational selection effect", and makes it impossible to say anything at all about the size distribution of terrestrial planets. However, most formation mechanisms of the type we generally imagine for terrestrial planets yield a distribution of more smaller examples than larger examples, with a lower cutoff at some "minimum size" for the described process to be in effect. Unless the Earth is already at or near that minimum size cutoff, I find it very unlikely that there are not vastly more smaller terrestrial planets than larger (notwithstanding the fact that the definition of "planet" may be pretty useless in this context).