The next technical step is actually a reworking of the template/lua code for producing that infobox.  The current one works for the few genes I've manually tested it on, but the way it uses arbitrary access (via nesting structures in the template code) it fails rather dramatically when data that it is expecting is missing.  Its also really inefficient.  See (and contribute to) thread about this on the module:wikidata talk page [1].

The step after that one could end up being much harder.  Converting all of the current infoboxes to use the 'one template to rule them all' is not technically very hard, but not everyone on EN Wikipedia is necessarily excited about this transition.  See e.g. the post complaining about our experiments by one very experienced Wikipedian [2].  The real sticking point is the challenge of making it easy for would-be editors of the data in that infobox to edit it.  Especially given the complexity of our data model here, its not exactly easy for a new person to find the right wikidata items to edit.  I'm hopeful that ongoing work will ease that process!

Happy pre-birthday Wikidata!

-Ben

[1] https://en.wikipedia.org/wiki/Module_talk:Wikidata#Efficient_pattern_for_gathering_many_attributes_from_wikidata.3F

[2] https://en.wikipedia.org/wiki/User_talk:JohnBlackburne#Gene_Articles_ARF6_and_RREB1