Bootstrap values usually give you useful information, be they low or high. Usually, a "high" bootstrap value is something >= 70%. That means a clade is strongly supported; lower values are telling you the clade is not clearly separated from others. This could be for several reasons, e.g.:
1) A subpopulation is not really separated from others, e.g. the viral sequences in town "A" and town "B" are completely admixed, meaning the population is panmictic and there is continuous gene flow between the two of them;
2) the populations diverged not long ago, it's too soon to have any phylogenetic signal, at least using the marker you are using (different genes evolve at different rates). Usually, housekeeping genes (which are usually very conserved) are used for populations that supposedly diverged long time ago.
The interpretation of your tree should take into account the biology and epidemiology of your target organism. If you really want to learn more, I suggest a lot of reading, and definitely a class dedicated to molecular phylogenetics.
This might be a good start:
Also, the NJ algorithm should allow for uneven branch lengths, maybe the tree you are showing is the consensus tree?
Hope it helps