Wyrd isn't going to be making any decision based on data alone. They are watching tournament results, community discussions, and playing games themselves. These are all factors, data is just another one for them to consider.
If you're concerned your data is sending the wrong impression though, I would encourage you to make the completed game state in your phone match what you expect the final result to be. Not hard to do that point-wise, even if you don't mark down all the model deaths and stuff.
I'm not convinced there's much statistical significance in the difference of silly vs competitive games.
But I think what the data is most helpful for, should they chose to use it, is seeing how often models are hired and how long they generally last (for games that mark model deaths). That and seeing how often points/strategies are selected/scored, which should even out among all the games since they should be equally prevalent in social/competitive matches.