FWIW I've filled in the survey. With the number of users of this plugin I dont think we are in any danger of the 30 day developer key "unused" scenario, so I would expect it to be usable up until the (unannounced) date of actual API retirement.
Personally I have no desire to attempt to replace the functionality with developing an html scraping based approach, that is just opening yourself up to a world of pain of the plugin breaking every time the website gets changed, as happens with all the metadata plugins. Not to mention the fact that while scraping works ok for reading data it becomes a lot more complicated/problematic for updating (posting) data given all the funky stuff people often do with javascript etc. Of course if someone else wants to create a new plugin to do this then good luck and I hope it works
. Goodreads have been particularly poor/lazy at adding to their website over the years (it has barely changed) so it does have that in it's favour.
So unless Goodreads have a change of heart I expect this plugin to die. I still use it every time I finish a book (my usage is to simply add the book to a yearly shelf in goodreads with my rating, and tick a #Read boolean custom column in calibre also setting #LastRead date column. So its just going to be a bunch more mouseclicks each time - to do the same thing - not the end of the world given I only read a few hundred books a year but a shame nonetheless.