Cameras obey the laws of physics. In this case the specific phenomenon is called diffraction, and the diffraction limit applies here.
For example, the 5x telephoto lens of Samsung Galaxy Note20 Ultra carries the following specs:
- Resolution: 12 MP (4000 x 3000, I guess)
- Equivalent focal length: 120 mm
- Aperture number: f/3
- Pixel size: 1.0 um
The physical size of the sensor element is 4.0 x 3.0 mm (5.0 mm diagonal by Mr. Pythagoras). This enables us to calculate the real focal length, which is 5.0 mm / 43.3 mm x 120 mm = 13.9 mm (43.3 mm is the full frame diagonal to which the "equivalent" refers).
Now that we have some numbers, let's get back to the diffraction limit. A point light source (infinitely small) correctly focused produces a non-pointlike spot on the camera ("Airy disc"). The size of this spot is roughly d = 2 x 1.22 x wavelength x (aperture number), in this case 2 x 1.22 x 0.5 um x 3 = 3.7 um for green light. Smallest spot of light on the sensor is thus almost 4x4 pixels! The image on the sensor is inevitably soft due to the laws of physics, even though the camera signal processing works really hard to make it look sharper.
Why not make the aperture number smaller? Well... The light accepting area of the lens is determined by the focal length and aperture number. In this case 13.9 mm / 3 = 4.0 mm, which is actually a very large lens for a mobile phone.
How about making the focal length longer? That would increase the relative magnification. Unfortunately, we'd still bump our head into diffraction. Doubling the focal length without changing the lens diameter doubles the aperture number. Then we would have even softer image on our camera element, and the only thing we would have achieved is smaller image area.
Larger sensor? No help here. It turns out that if the physical aperture of the lens is kept constant, the maximum angular resolution is the same. For a 4-mm lens the angular resolution is approximately 1.22 x 0.5 um / 4 mm ≈ 0.00015 rad ≈ 0.0088° ≈ 0.5'. (By the way, normal visual acuity for humans is 1.0', and some individuals may achieve 0.5' resolution.) So, the system is diffraction limited, and that is the end of the road.
I doubt 10x would bring any significant improvement even with today's AI image processing, but it would certainly limit the image area. Unless, of course, someone can really make the lens bigger and still squeeze the optical path into a mobile phone. Not easy.
And just for reference: I took the specifications of an inexpensive (one quarter of an iPhone 12PM) superzoom compact camera, Panasonic FZ82. Its maximum focal length is 215 mm (1200 mm equivalent), and the aperture number is 5.9 at that zoom. Thus, the physical aperture is 215 mm / 5.9 ≈ 36 mm and the diffraction-limited resolution would be 0.06'. Almost tenfold aperture (36 mm vs 4 mm), and tenfold zoom reach (1200 mm equivalent vs. 120 mm equivalent).
(Yes, I am cutting a few corners here. I suspect the lens used by Panasonic is not quite diffraction limited, but by the sample images it is not much worse.)
