Yes, sensitivity is tied to focal length.
You can measure the focal length with pixels, which is a physical measurement since they also have a physical size. So different monitors will need different field of views to end up with the same focal length and sensitivity.
The focal length is basically the radius for the spherical image. A high focal length will be a large radius, resulting in a large sphere, and since the monitor size is fixed, it will result in a smaller portion of that sphere fitting inside the monitor space, resulting in a lower fov and an image that has a lower curvature. A shorter focal length is the opposite. Curvature plays a huge role in how the sensitivity feels, which is why even a perfect conversion (0%) to a different focal length still feels different. 2D would be kind of equivalent to an infinite focal length, as it is an image with no curvature.
The ratio between the spherical representation of the sensitivity (measure as cm/360 degrees), and the spherical representation of the image, is the actual sensitivity, the 'Control-Display Ratio'. This is often expressed as 'Control-Display Gain', which is how much faster or slower the output is compared to the input. This actual sensitivity is a constant value as long as you convert using 0%.
You can find your actual sensitivity using this. The defaults in the calculation is a result of doing a desktop to csgo conversion using 0% for a 24.5" monitor at 1920x1080 and 400 CPI, which results in an identical 2D and 3D Control-Display Gain of ~4.45. Someone may think to themselves that a 1:1 Control-Display Ratio would be optimal, but it will feel incredibly slow, even for just 2D, which is probably due to the mouse mass, friction, etc.
Also here are some useful visuals and links:
Graphical FOV Converter
Spherical representation of sensitivity and image at high and low fov.
Small monitor, small fov, large monitor, large fov. Same focal length. Same cm/360°.
Same fov, different monitor size. Different focal length. Different cm/360°.