If you make only a half step to learn internals about the C compiler and the linker, why not doing the full step and write C at all ???

How is a published operation (OVERLAYABLE) an 'internal'
just do nut publish as a C adaqvntage something that is available in assembler as well.

I do "write C at all" sometimes with a few assembly modules.

A blatant example of the assembly advantage:

you have a block of <=256 bytes on XDATA that you want to process and move the result to another location in Xdata.
Using whatever method you prefer, you locate that block within in a page of memory (same high byte of address for all).
NO C compiler (because C refuses to be concerned with locations) will set P2 (or the high address SFR) ONCE and use movx @Ri

Now with "blatant example of the assembly advantage", that does NOT mean that there always is a noticeable advantage to using assembler. I use C as much as timing allow. I would even add that if you need assembler to squeeze 2% out in code size ot time, you are better off using a faster processor. However the above example improves process time more than 50%.

Erik