I'm not a physicist but I do a lot of recording. As I understand it you record the frequency that is there and any frequency that is absent cannot be added by way of equalization. However, the actual volume of the individual frequencies in a recording depends on other factors. This may be equalization or lack of at the engineering stage, compression, digital conversion, microphones, equipment and so forth. If a frequency exists in a recording it can be boosted, lowered or cut.
So to look at a spectrograph and say 'this violin has a low reading at 200hz' may be telling you more about the recording. I'm not surprised a Luis and Clarke violin sounds weaker at the low end than all of the professionally recorded examples you give seeing as it was recorded on a video camera probably with severe compression.