WebODF simply loads the ODF xml as is and adds CSS to make it render well. This means that no information is lost. The same approach does not work well with OOX

WebODF simply loads the ODF xml as is and adds CSS to make it render well. This means that no information is lost. The same approach does not work well with OOXML which makes it harder to write a similar library for those formats. For the binary formats the same approach is completely out of the question. One way to build similar functionality for those formats is to take existing ODF converter code from Calligra or LibreOffice and port it to compile nicely with emscripten. The result can then be edited with webodf.