Fixes #7.

This ended up opening several large stinky cans of worms. There were a lot of assumptions in the current code around immediate CLI output that needed changing. Additionally, displaying information in a GUI requires more discrete access to data than the current program allowed. (i.e. no easy way to get the currently assigned probe level).

The end result was perhaps more refactoring than I planned on (or perhaps you wanted...). I had to touch the important bits today (the probe arrangement algorithm) and while it doesn't seem like I broke anything, I wanted to start a PR to get some more eyes.

Big changes

• Requires Qt >= 6.7
• Bumped to C++20
• Added vcpkg to bring in deps. Can still use system deps without vcpkg, but makes it easier.
• Use spdlog for command-line output, such that when run as a CLI it outputs to console as before, but when run as a GUI output is logged.
• Site/Probe properties are compiled in, rather than be loaded from file at runtime. This was needed to give users a starting point; the alternative was forcing a text editor which, I feel, would defeat the purpose of a GUI.

Things still on my to-do list:

• Recolor site links if they are providing a benefit.
• Allow manual changing of assigned probes.
• Display real-time changes to totals instead of only after running a solution.

One stray observation - while the program will build and run on Windows (VS2022), the simulation does not. I check the current code on Windows, and it freezes on the first iteration as well. I suspect multithreading issues, but haven't poked that hornets nest yet.