Field homogeneity is critical for performance of many multi-pulse NMR experiments.

Taking your example of a sample where the natural linewidth is ~8 Hz, then it is likely that it is a biomacromolecule. I assume that the large linewidth of the NMR signal in your sample is due to size of the molecule and not due to chemical exchange. HSQC Spectra of proteins are usually acquired in water. Although the intensity of the protein signal does not change much due to small magnetic field inhomogeneities (as you mentioned), the LINESHAPE of all signals is affected. The most important consequence is that the water signal which (before suppression) is many orders of magnitude larger than the protein signals is more difficult to suppress by using presaturation (even small inhomogeneities can lead to distortions of lineshape) resulting in a large visible hump even after presaturation. These artifacts are especially severe if the magnetic field inhomogeneities lead to non-lorentzian lineshapes. These humps not only distort the baseline but also vary from scan to scan resulting in substantially higher noise. In addition the much larger residual water signal necessitates a reduction in the receiver gain which again leads to loss of sensitivity. Use of pulsed field gradients is especially helpful to reduce the magnitude of these artifacts.