Harmel et al. [Appl. Opt. 50, 5842 (2011)] presented an intercomparison of products from collocated above-water radiometric measurements performed with a Hyperspectral Surface Acquisition System (HyperSAS) and a multispectral Sea-viewing Wide Field of View Sensor Photometer Revision for Incident Surface Measurements (SeaPRISM). Radiometric products from HyperSAS data were determined with a processing code written by the authors, while products from SeaPRISM measurements were obtained with the standard processor of the Ocean Color component of the Aerosol Robotic Network (AERONET-OC). Despite the application of equivalent processing schemes, results exhibit spectrally averaged unbiased relative differences of $+26\%$ between HyperSAS and SeaPRISM exact normalized water leaving radiances, $L_{\mathrm{WN}}$. Harmel et al. concluded that the sun-glint correction scheme applied for SeaPRISM data reduction is a major reason for the observed differences. This comment rejects the former conclusion as being supported by a wrong interpretation of the AERONET-OC processing scheme, and a consequent failure in describing the spectral properties of the glint radiance determined from SeaPRISM data. Afterward, the differences between HyperSAS and SeaPRISM $L_{\mathrm{WN}}$ reported by Harmel et al. are challenged with intercomparison results from collocated measurements periodically performed over almost a decade with an in-water multispectral system and SeaPRISM. Results for $L_{\mathrm{WN}}$ from the in-water system show spectrally averaged unbiased relative differences of $+1\%$ with respect to SeaPRISM values.