Yes, this is definitely doable.
It should be possible to use PyOxidizer's build system to construct a static or shared library which you could then consume from your C/C++ application's build system. Simply invoke e.g. `pyoxidizer build-library` to produce a library and add e.g. `-lembeddedpython` to your application's linker arguments and things should "just work." Alternatively, you could hack up the generated Cargo.toml to produce a library instead of an application and then hook that library up to your C/C++ application's build system. Alternatively, you could configure the Rust application to link your C/C++ application and then call into its main(). This would have Rust/Cargo handle all the linking bits. There's tons of options. All of them probably doable today without modifying PyOxidizer's source code. But we should get something documented so it is minimal effort for others.
The only real complexity here is with loading modules from memory. That requires custom Rust code that isn't part of the Python distribution itself. If you want to fully embed a Python interpreter into a C/C++ application, you could consume the libpython static library in the Python distributions used by PyOxidizer. To get in memory import will require executing some Rust code in addition to what's in the Python distribution.