Ok, either way is possible, it’s just a bit more complicated if QGC and MAVLink are required 
Nope - it just means that the steps for how to do so aren’t already clearly written down in a single convenient place.
That's the approach we're working to enable ...
Unfortunately I haven’t yet had the chance to properly document the extension creation process, because our extensions infrastructure is new, and there’s some other documentation I’m trying to finalise first.
We do have an examples repository, which is intended to be a set of steps of progressively more complex extension features. We’re planning to have a persistent logging example, but in the interim it’s hopefully helpful to refer to example 4 (which includes persistent file saving).
What kind of timeline are you working on?
Basic sensor integration is something we expect to be a reasonably common use-case, so is definitely something we could benefit from having a simple example for, and we’d be happy to work with you to make sure you’re able to create that if you want to.
That said, the fastest and simplest way to get data when starting out will likely be:
- Use the USB over IP extension discussed here
- Installing
pylibftdi(with pip) on your topside computer (you’ll need Python first, if you don’t already have it) - Write and run your Python logging code there (making use of the Atlas Scientific examples, starting from the “testing installation” step in the README).
USB over IP should enable minimal “data retrieval” functionality from a script running on the topside, but means you need to manually start your program each time you want to start collecting data. If you want the convenience of your code automatically running on the vehicle, when the vehicle turns on, and/or additional features like having a graphical interface and being easily installable on other vehicles then we can work together to turn your script into an extension