I received a link to this poll in my e-mail through the Entomological Collections Network listserv. While I appreciate being able to give my imput, most of the important questions, (e.g. "do you think molecular techniques help in species discovery" were yes/no questions with no space for comments. The most interesting of these was "Do you think molecular techniques can accelerate species description?" To which I answered no, emphatically.
Despite complaints to the contrary, species description is not a particularly difficult process. While Botanical and Zoological Nomenclature differ in that the botanical code requires the primary description to be written in Latin, they otherwise have all the same features. There is a brief introduction and materials and methods, followed by the diagnosis (short description meant to separate the new species from other known members of that group) and description (longer statement of all aspects of the morphology). Typically there are illustrations, which can be achieved with a microscope that has a grid eyepiece (or a camera, if you're lucky enough to work with larger organisms). Then the material examined, remarks on biology, systematics and distribution, and works cited. I have already discussed techniques to accelerate the activities that go into species description before the actual writing part. Notice that none of those include any molecular techniques.
The simple truth is, while molecular techniques may speed up the process of resolving cryptic species complexes, they seldom (if ever?) increase the speed at which species are described. Current estimates are that there are millions of undescribed species on Earth, most of those being arthropods. And I propose that many of these are sitting in museums and unsorted trap samples. The ICZN still requires a formal description for a name to be recognized, and without a name it's very difficult to talk about something. If you look at what characters people are using in their descriptions and diagnoses, 99% of the time these are gross morphology. And, for the vast majority of all the species yet to be described, this is all that is necessary. To make things easier, there are journals such as Zootaxa which specialize in publishing descriptions of new species, and have accelerated peer review and publication time within a month.
Accelerated species description will not occur when everyone has a lab equipped with a pyrosequencer. Accelerated species description will occur when biologists as morphologists spend more time contemplating Evenhuis' Steps to Enlightenment and Taxonomic Nirvanna. The only way species get described is by writing a paper, and the only way diagnoses and descriptions get written is by looking at the organisms. Collecting from poorly sampled regions, sorting old trap samples, building taxonomic libraries, photographing and illustrating type specimens, and many many hours at the microscope examining morphology, these are how species description will accelerate. Not by molecular techniques.