Yes I think this could be correct, from the code it is unclear if the vpp is used in DoVi or replaced with another approach,
So potentially:
vd1 :: p8 file HEVC YCbCr 4:2:0 Storage → convert to YCbCr 4:2:2 Transmission using bicubic → viu (csc to ICrCp 12bit) → vpp
osd1 :: graphics → viu (csc to ICrCp 12bit) → vpp
vpp:: blend → add scrambled RPU + CRC into chroma lsb → send data inside RGB 8 bit 4:4:4 (tunnel).