Nothing new here, just some additional scripts which help me get from ODF files (primarily ODT files) to ePUB 3.0.1 files which pass the epubchecker validation with zero errors.
Does not use either Calibre or Sigil. Calibre is fine for managing ePUBs as a consumer and getting them on and off devices, but not as a publishing tool. Sigil is fine for ePUB 2.0.1 and earlier, but does not meet the requirements of the current standard.
Current method relies on the DITA Open Toolkit, the DITA for Publishers specialisation and the ePub plugin. I use a modified version of oXygenXML Editor to tie all this together. It's not free, but it works and it works well.