That's basically what's going on, but it's all a manual process so far. The goal is scoped really just to the actual georeferencing and mosaic creation, no post-processing or OCR is performed on the maps within this platform (people are certainly downloading the final rasters and doing that with them though).