Time spent with tpis not wasted. It is clearly the most successful small language so you can learn a lot about what works (in general, not particular items, perhaps). On the other hand, it has many flaws (often in particular, but also in general) so that you can get an idea what to avoid. One thing that springs to mind because of your work here is the need to get some system to the vocabulary, which is pretty damned random in tp and leaves great gaps in practical and experiential areas. On the other hand, the totally systematic languages have mainly died aborning, so some variety is called for. Good concept maps (distributional semantics) for any language might well be useful.