GeoTiffs are images containing geolocational information. Although, tiffs are supported, the geo information is not recognized by RC. You could align the image manually via setting the x,y,z coordinates in the space, but it would be a dumb plane at that region.
GeoTiffs can contain - when used as DEM representations (Digital Elevetion Models) - 3 dimensional information. The intensity of the pixel describes the height of the landmark in that spot.
This data is commonly available for most regions in Europe and could vastly stabilize the algorithms of Reality Capture, especially for urban areas.
Both perspectives in the screenshot show the same single GeoTif image. This is NOT a mesh or a volumetric model!
Correlator3D for example is able to interpret this data, but has no smart feature extraction algorithm.
Reality Capture has smart feature extraction and a good feature matching algorithm, but no capability to add DEM information.
I am really convinced, you should think about integrating this feature to combine the power of both world.