I think this question has no definite answer yet because the evidence is not enough. Our knowledge around the viral quasispecies is proportional to the power of tools and methods we have used to study the viral quasispecies. For instance the sensitivity of sanger sequencing for detection of a unique nucleotide sequence of virus is 20-30 percent that means that virus should comprises 20-30 percent of the viral population. The sensitivity of the next generation sequencing (NGS) is around 0.1-1 (0.1 is very optimistic!) percent which is limited by the error of the polymerase we use in the PCR amplification. As a result, with the updated tools and methods such as NGS just we can find a virus with a unique nucleotide sequence which comprises 0.1 percent (10E-4) of the background viral population (currently 0.1 percent is our limitation). Please note that when we use NGS, mostly we find more polymorphisms in the viral population in comparison to when we use sanger sequencing which shows that we are more successful in studying viral quasispecies using NGS and consequently, if we investigate with more powerful tools, we would find a wider population of the virus.