Ideally, visible BSDF measurements are done with a broadband source and a detector with a filter that multiplies to a photopic response V(lambda) when combined with the source spectrum. This is then a CIE-Y channel measurement.
A monochromatic source cannot deliver an equivalent response, and your measurements will be assuming the surface has little or no coloration in that case. You are effectively using the single wavelength measurement as a proxy for a proper spectral sensitivity.
For near-infrared light, I do not know the standard practice. Someone else will have to answer this question.
You can find descriptions of the light sources used in various instruments in the recently published IEA Task 61 C2.2 publication. For example, pabopto pgII uses a tungsten halogen light source with filtered Si and InGaAs detectors for VIS and NIR wavelengths, respectively, at LBNL. HSLU (Lars Grobe) uses the pgII with a halogen source for VIS and a Xenon arc lamp for solar.