First I have to say that the wavelengths of light are on a totally different order of magnitude than sound. So the parallel drawn in your question "do light waves, for example one with the same wave length as a mid-C and another with a mid-F wave, look nicely together?" may seem logical, but is on closer inspection not easily maintained. Instead, one way to address this question more appropriately would be to talk about wavelength differences; i.e., light with wavelengths differing X octaves compared to acoustic tones differing X octaves.
Having said that I think it is worthwhile to sidestep the theoretical approach and take a closer look at how auditory and visual sensory information is actually processed at a neurophysiological level.
Sound is processed in the inner ear (the cochlea), which basically works as a Fourier transformer, and specifically a frequency-to-place converter. The spatial distribution of the characteristic frequencies (the tonotopy) on the basilar membrane of the cochlea follows a pattern where one octave spans about 2.5 mm (Greenwood, 1990). For the approximately 10 octaves the human ear can hear (~20 Hz - 20 kHz) we have 16,000 inner hair cells. See the following picture for a rolled-out cochlea with the tonotopy illustrated:
The eye, on the other hand, analyzes light frequencies using just 3 colors (red, green, blue) by means of three cone classes (as opposed to 16,000 hair cells each sensitive to a slightly different acoustic frequency). Although the visual system does a great job in combining this sparse frequency information into a spectrum of colors, it is not a frequency analyzer as such. In fact, it is more of a frequency "combiner", as it combines the ratios of activation of the cone classes and runs it through a system of color opponency. By weighing the relative contributions of the three colors (RGB) using the opponent system (R-G, Y-B) the visual system estimates the color of the object you are looking at. Below on the left is illustrated the spectral frequency sensitivity of the three cone classes (note the unequal, non-octave distribution of the three across the dynamic spectrum range), and on the right the color-opponent (Hering) model.
source:huevaluechroma.com and giantitp.com
The opponent nature of human vision (blue-yellow and red-green axis) results in a 2-dimensional color space which is very different from the 1-dimensional frequency space of the cochlea (Mather, 2006). Note that the third visual dimension is brightness, comparable to the second dimension of loudness in the ear.
In all, based on a neurophysiological signal-processing point-of-view, hearing and seeing frequencies are two completely different things. Comparing octaves between the two is worse than comparing apples and oranges, as apples and oranges share at least the same dimensionality.
It probably doesn't answer the question, but it may answer the question why it cannot be answered in any logical, comprehensive and straightforward way without loosing oneself in subjective monologues about "I like this and this combination of timbres but I don't like this color combination so much" kind-of-thing. Admittedly, it can be experimentally addressed by inquiring about the subjective 'pleasantness' of a set of combinations of colors and pitches in a study population. But from a physical and pysiological perspective, comparing the two entities of light and sound through equal-octave comparisons doesn't make sense since: (1) the two entities are processed through completely different neurophysiological principles as described above, ('1D acoustic frequency Fourier analyzer' versus '2D spectral combiner'), and (2) they represent entirely different physical entities altogether (photons/EM waves versus air pressure differences). Their only commonality is that they have an oscillatory nature, but that's where the parallel starts, and ends.
Greenwood et al. JASA 1990; 87:2592-605
Mather. Foundations of Perception 2006