In short Limusaurus seems to be unusual not just because it is a basal herbivorous theropod, but because it is so profoundly different to the other members of its clade (the ceratosaurs) that are so obviously carnivorous. The other theropod lineages that yield herbivorous / omnivorous taxa are pretty uniform in their dietary approach so this appears to be a quite an exception.
Now for the caveats to this. First of all, was Limusaurus really a herbivore, or possibly some kind of omnivore or even a carnivore? Well, obviously we think the former, and while omnivory is a possibility (and here I mean genuine omnivory of eating lots of different foodstuffs and not just the very occasional egg or insect which is common for many herbivores) carnivory can be largely ruled out. In either case, the profound difference between Limusaurus and other ceratosaurs would still stand out both in terms of diet and feeding apparatus (and indeed overall body size).
We can turn now to the other theropod lineages that could be considered herbivores for comparison. First off the therizinosaurs (such as Beipiaosaurus), which are relatively easy to deal with. These all have skulls and teeth that are ‘classically’ herbivorous, with leaf-shaped teeth useful for cropping leaves and little else and there is no real dissent to the idea that these were herbivorous animals. Importantly to the point above, they are all pretty similar to each other as well – there’s little variation between taxa.
Next the rather more complicated ornithomimosaurs. The diet of this clade has been the subject of much discussion, but I think it would be fair to say that they are generally considered to have likely been herbivorous though both omnivoroy and carnivory have been strongly supported by different authors. There is more variation here in the feeding apparatus (compare the very many small teeth of Pelicanimimus say to the beak of Gallimimus), but again in general the morphology is conservative. Even if there is variation in the diet, this is likely to be part of a continuum of diets based on limited morphological variation and not the disjunct diet / skull of Limusaurus compared to other ceratosaurs.
Finally, and most awkwardly, there are the oviraptorosaurs. Again the diet for this clade is controversial and runs from herbivory to carnivory (including, of course, egg eating) depending on which taxon and what evidence you are citing. However like the ornithomimids, while there is some variation in the structure of the feeding apparatus (e.g. compare the oddly toothed Incisivosaurus with the beaks of Caudipteryx or Citipati) this remains generally conservative and if there is significant variation in the diet between say the taxa listed above, again it will likely be as part of general continuum and not a big separation between well toothed and gastrolith-ed herbivores, and sharply beaked predators.In short then, the ceratosaurs do appear to be unusual in having Limusaurus in their ranks, because it is apparently so different in size, morphology and diet when compared to the other taxa in the clade. There is, at the moment, a great discrepancy between Limusaurus and its relatives and one that is not likely to be breached even with extensive new fossil finds in that even with a continuum from strict carnivory in say Ceratosaurus through to strict herbivory in Limusaurus, these animals at the two ends of the spectrum will be far more different in morphology than the most disjunct oviraptorosaurs or ornithomimosaurs (again that we have so far, basal members of those clades may alter is picture significantly).
This is, really, a pretty minor point when it comes down to it – the ceratosaurs may be more diverse in terms of feeding ecology / morphology than other theropod clades. However, the interest that understandably lies with theropod ecology and the apparent frequency with which some theropods turned their backs on animalian food is interesting. In the ceratosaurs (*potentially*) we have an example where only one (or more likely a few when one includes possible close relatives like Elaphrosaurus) animal went down the route of herbivory while the rest of the clade stuck to carnivory, when in other theropod clades where herbivory has apparently evolved, what is true for one may well be true for all (or at least part of a close association of similar diets). It’s convoluted, complex and full of caveats, but potentially quite interesting and something I’d hope to try and explore further in the future.