Edit: I think I found your camera, PowerShot SX70 HS. It apparently allows remote zoom control and is equipped with a whopping 65x zoom.
The best camera for your situation, because you don't absolutely need a large sensor and can replace a large sensor with long exposure, is a compact camera. Unfortunately, Canon has discontinued programmatic API control of its PowerShot cameras, as you have noticed. Some old PowerShot and IXUS cameras can be controlled remotely: http://www.gphoto.org/doc/remote/ ...but you need to find a used camera and its Mpix rating will be poor. So, if you cannot find any recent remote-controllable compact camera, your only option is a DSLR. Compact cameras have the advantage that the zoom happens with an electric motor, meaning it's possible in some cases to control the zoom programmatically.
EOS DSLRs offer one advantage: bigger sensor (which you don't need) and one big honking disadvantage: the zoom is not controlled by an electric motor, so zooming can only happen manually. So, if you decide to go with EOS DSLR, you need in practice at least two, perhaps even three (and slight amount of digital cropping).
The Canon camera API supports EOS 4000D which is usually shipped with a 18-55mm non-image-stabilized zoom lens. You could purchase two, for $300-$350 a piece, and set their zooms to different values. 4000D is 18 Mpix, but if you need to crop the images, it will become smaller. Do note the optical quality of 4000D and the shipped lens is quite good, and noise is low, so the 18 Mpix is 18 true megapixels, not some useless advertising value.
If you want to go a bit more expensive, EOS 2000D supports 24 megapixels and is usually shipped with an image-stabilized lens. Are you willing to pay for the extra 6 megapixels, only you know. The image stabilization in your case is an unneeded feature.
2000D and 4000D can also be placed easily on a tripod using a standard screw mount.
Your zoom need is 6x. If you set one lens to 2.45x the other, in worst case you need 2.45x digital crop in both directions. 18 Mpix thus would in worst case be 3 Mpix (18 / 2.45^2 = approximately 3). But the images at 3 Mpix would be extremely sharp, so perhaps that could be enough for you? 24 Mpix would in worst case be 4 Mpix.
If you purchase three DSLRs, and set:
- One to 18mm
- One to 31.5mm
- One to 55mm
You in worst case need 1.96x crop, meaning 18 Mpix turns to 4.7 Mpix and 24 Mpix turns to 6.2 Mpix. If you can move the 55mm camera a bit closer, you
- One to 18mm
- One to 32.7mm
- One to 55mm and move a bit closer
...you in worst case need 1.82x crop, meaning 18 Mpix turns to 5.4 Mpix and 24 Mpix turns to 7.2 Mpix.
My advice? If you plan to use the cameras for other purposes as well in the future and not just for this project, buy one 2000D with an image-stabilized lens and learn how much/little megapixels mean. Then you can decide which one/two other cameras to purchase, 2000D or 4000D. Perhaps even 3 Mpix would be enough for you if the images are really sharp?