Improve Q&A info navigation by understanding R and SQL logical conventions

Currently Q&A natural language only works for the simplest data statements. For almost all of our data models, it would help A LOT if Q&A could understand some specific logical statements which would allow us to identify non-ambiguously some subsetted information to chart. I can add examples if you'd like. Currently Q&A is too primitive for more complex data models with some ambiguities.
(although I love the natural language dream, it's not really good enough yet for our scenarios and data models.