I don't think that is the main issue. Its not THAT hard to build a secondary sqlite index for data stored in git and keep in in sync especially because the data is already versioned and can be incrementally updated.