@Oswald: agreed, only the sample holder should be different and the spectrum of the substrate should be taken into account (you may use pure substrate as comaprison in the second ray path - only it has to be of identical thickness)
True. you always need a reference. I often fixed my samples with some tape just above th hole to the detector. Sample holders are overrated for such questions.
If you are looking for optical transparency then yes, a transmission / absorption spectra should give you the transparency of the films, either free standing or on a substrate that is transparent within the wavelength region of interest. As Gyorgy and others mentioned you must take into account the effect of the substrate. However, quantitative treatments on solutions and thin films are very different. I suggest you look up -transmission in multilayer thin film- in google.
Ofcourse this does not help you if you are not looking for "transparency in the optical regime".