Fix GH CI size-diff by switching from make to meson

[ALTracer]

Detailed description

• No new features.
• CI size-diff is failing because it relies on effectively make PROBE_HOST=native which has been overflowing flash for some time.
• This PR solves it by switching to building via meson, which since meson: Disabled some of the default targets to make the code fit again. #1964 should fit (and is controlled more easily).

I wish there was a way to also report size-diff for 256-512 KiB probes with every driver compiled (and every applicable log statement compiled), but let's unblock the existing workflow first.
Tested on own fork by manulally triggering CI. Caveat: I asked for meson compile -C build-native bin the firmware bin alias, so it runs faster (omits BMDA), but still downloads hidapi/libusb/ftdi submodules for some reason. Network-isolated containers should not access internet past "install bdepends" and "fetch sources" stage (Gentoo ebuild FEATURE=network-sandbox).

Your checklist for this pull request

• I've read the Code of Conduct
• I've read the guidelines for contributing to this repository
• It builds for hardware native (see Building the firmware)
• It builds as BMDA (see Building the BMDA)
• I've tested it to the best of my ability
• My commit messages provide a useful short description of what the commits do

Closing issues

[dragonmux]

The PR was much appreciated, however Esden finished the build system PR (#1967) which also fixes size-diff. We'll close this as its purpose has been served.