Welcome to Rangefinder Forum.
It's of much help to know which frames were shot at which speeds. If a focal plane shutter such as used in the Leica screw mounts needs servicing it will often be the fastest speeds and/or the slow speeds which will be most affected. Slow speeds because the escapement may benefit from cleaning, the fast speeds because the higher the speeds are the tighter the slit width needed becomes, and the tighter this becomes the less tolerance exists for even minute variations in curtain travel of a millimetre or so.
Looking at the colour roll it seems the full frame has not received exposure. Unless you possibly have been inadvertently fouling the shutter speed dial with a finger it is hard to envisage any other cause than an irregular shutter which, if the camera has not been serviced for some years is not at all unlikely. I don't see how a lab could develop a portion of certain negative frames adequately without also achieving identical development of the remainder.
I would therefore suggest that the most badly affected frames were likely those exposed at 1/1000 to perhaps 1/250, certainly 1/1000 or 1/500 as it seems the shutter is capping early in some cases. It is entirely feasible for a IIIf in need of service to produce satisfactory exposures at 1/60 but not at its fastest times.
If you would like to be definitive load a short roll of film (black and white is fine or even preferable, given you can control the entire process) and take a series of equivalent exposures at speeds from 1/1000 down. Ie different shutter speeds and adjusting aperture to maintain same exposure value. A landscape with plenty of cloudless evenly lit sky is ideal, but a well illuminated light coloured wall or building that is evenly lit will also work. Obviously you will need to note the exact exposure settings for each frame. On developing the roll examination of the various frames will inform what the shutter is doing at each speed. If you post these we can help assess them.
With most cameras a decent shutter tester will reveal any shutter problems quickly and easily without the need for a test roll, unfortunately with the IIIf film gate being inaccessible to most tester sensor probes, the old fashioned method is usually the easiest unless one is prepared to remove the mechanism from the body casing.