The concept of punctuated evolution – bursts of evolutionary novelty separated by long periods of stasis – was first proposed by Stephen Jay Gould & Niles Eldredge in 1972. Since then, there's been an ongoing debate among evolutionary biologists about how significant ‘punk eek' could be in the evolution of new species. (Remember that they aren't arguing about evolution itself, but about the relative speed of evolutionary change.)
Mark Pagel and his co-workers recently published a study of divergence in gene sequences (Pagel et al. 2006). Their data indicated that around 20% of genetic divergence could be put down to rapid, and not gradual, change, and they concluded that "[p]unctuational episodes of evolution may play a larger role in promoting evolutionary divergence than has previously been appreciated" (ibid.). What could this imply for speciation?
The researchers suggest that these underlying genetic differences could reflect a relatively high frequency of sympatric evolution. Some scientists feel that sympatric evolution, far from being a rare event, is actually quite common. For a new species to evolve within the range of the original, parent, species would require the rapid appearance of reproductive isolating mechanisms separating the two. These could include things like mutations affecting signalling molecules, behavioural change (e.g. indigobirds), changes in pollinators, and physiological changes that make gametes incompatible – all of which might correlate with rapid genetic divergence. Polploidy, which is far more common in plants than in animals, is also likely to produce small daughter populations that can't mate with the parent population. (And the frequency of polyploidy in plants may explain why examples of sympatric speciation are more common in plants than in animal groups.)
References: M. Pagel, C. Venditti, & A. Meade (2006) Large punctuational contribution of speciation to evolutionary divergence at the molecular level. Science 314: 119-121