hi, i'm finalizing a 2D ns solver on unstructured tri-grid with fancy interpolations allowing for use of low quality meshes (yes, I get perfectly horizontal flow in a tube on crippled triangles). do you think it's worth making it into a C++ class and publishing it in some technical paper? when starting I couldnt find anything relevant that I could just take... thanks

Most definitely, if a literature search shows no sign of the same scheme, you should publish it in a technical conference or journal paper. You should also investigate whether the scheme could handle only smooth solutions or does it also do a reasonable job resolving shock waves on low-quality triangular meshes.