The 'modern lens grinding technique' is of course balderdash, but underneath is the fact that most if not all camera lenses can cover a larger spectral range than can the human visual system. For instance, IR-sensitive films and sensors used with the same lenses do form images.
So what we must point to instead is the resultant spectral coverage of the sensor and any built-in filtering applied by the coating on the sensor's cover glass. And just about all cameras designed for 'ordinary' photography will have IR rejection in the filtering if the sensor is sensitive to the near IR. Else the color balance could be harder to achieve, and chromatic aberrations could be unacceptable.
As an amateur astronomer, I'm well aware of such factors. Even to capture the red of hydrogen alpha emission, which is nearly 50nm blueward of the 700nm visual spectral limit, is often compromised by the built-in filtering. And so modified filters can be obtained for some cameras which have an extended red response whereby the red end cut-off is steeper (but not extended into the near IR) so that the H-alpha emission line is much less aggressively attenuated.
At any rate, consumer grade cameras have no better spectral coverage than our own eyes, unless specifically operating in a 'night shot' mode wherein there is extended spectral coverage into the near IR. Such a daytime photo as shown in the OP's linked-to page is almost certainly not covering outside the same ~400-700nm visual spectral range.
The presence of glass which is strongly indicated is seemingly providing a reflection, providentially positioned.