Surveys have always been of major importance in astronomy. Mapping a large fraction of the sky at an unknown or rather unexplored wavelength has always lead to serendipitous discoveries, although these discoveries should not be the main driving force for making a survey. Whenever astronomers looked at a new wavelength interval they discovered new objects or new phenonema. Looking at the sky in the optical reveals mostly nearby stars in our own Galaxy. A few fuzzy objects seen by eye are resolved into galaxies by large optical telescopes. When the radio frequencies were used in astronomy it turned out that most objects were not stars, but galaxies. And the sizes of these galaxies were much larger than their optical counterparts. When using infra-red telescopes you mostly look at young warm objects like starforming regions and young galaxies. At the highest energy part of the spectrum (X-rays and gamma rays) you see the most violent processes in astronomy, like possible black holes in galactic nuclei and also supernovae remnants.
One could say that the first thing an astronomer does when he is exploring the sky at a wavelength that has never been used before is making a survey to find out what is out there. Then he can take a better look at the most interesting sources afterwards.
A lot of modern telescopes have such a high resolution that it is not possible to carry out an all-sky survey . Because they don't have a large field of view, the amount of telescope time plus the amount of computing power to transform the raw observations into usable maps and catalogues is enormous.