In theory if the same cell is used and the solvent is the same for the Blank and the Sample solution then the L* for nearly colorless liquid should not be greater than the repeatability range of the instrument. Note, this also assumes no UV or Visible activated fluorescence or bioluminesce is occurring.
If the cell used to Blank is not the cell used to measure the sample then the user should experiment to determine if the cell is the cause of the difference. To do this you can standardize on Air, then measure a group of empty cells first, then cells filled with the same solvent and examine the range of readings.
Quartz cells are typically only required when wavelengths in the range of 190nm to 350nm will be measured. For HunterLab instruments Glass cells are sufficient for use, and plastic (PMMA) can be used if the user conducts a risk analysis as plastic cells typically have greater spectral variance than glass cells.
If the solvent for the blank is not the same as the solvent for the sample then this becomes a physics/chemistry question where refractive index, molecular composition, etc. can affect the readings.
For example it has been seen that when blanked on water, reading methanol based solutions L* values of 102 to 104 are common. Other organic solvents produce similar results relative to water blank. Normally most procedures specify which solvent to use when standardizing the instrument. Never assume that one should always use water as a blank, especially if the sample solution does not contain at least 50% or more water.