Hello Joan,
First of all, I would like to congratulate you for this magnificent piece of software. It is as good as the pro ones but freeware, which makes live easy for teachers like me. In fact, I am preparing some experiments to perform with my students.
Having said that, I must admit I am a newbie in the matter. Sorry if I am making a huge mistake in my demands.
I find particularly interest the perspective grid, recently introduced in the experimental version. I tested with real images and the results are quite accurate. Congratulations on that.
My aim at this point would be to go a bit further: once you have set the horizontal perspective so that any measure in that plane (ground floor typically) would become exact, a second calibration can be done in the vertical plane with an object which geometry is known (for example a wall). With these two calibrations, one for horizontal and one for vertical, we can obtain angles and distances from both horizontal and vertical objects.
Think of this example: you are analyzing the gait movement and you wanted to know the height of the knee. At this software stage, this can only be done for one plane, the plane of analysis, in which vertical distances match the real world. But if you have two runners at different distances from the camera, you need to choose between either of them because not all measures will be correct... unless you can calibrate the change in distance with this second perspective grid. I don't know if I have explained myself properly. 
Again, congratulations for this magnificent piece of software and for allowing the rest of us to use it without restrictions. People like you make most other people's day.
All the best for year 2014!
B.