No, your math is correct :-) : The only argument I see, is that the measurements you've done not precise enough, as it's hard to draw exact tangents for measuring the distances to the "Gartner sun". But that might be solved by using a scaled up image of the "Gartner Universe". Of course, the challenge remains to translate the significance of 1.3 (1.7)% difference of distance into meaning.
T 853 x^2 727609 785 y^2 616225 1638 1343834 SQRT 1159.23854 Q 914 x^2 835396 746 y^2 556516 1660 1391912 SQRT 1179.7932 T / Q -1.7%
Yes you could do that.There's a couple of information available how to do that (you probably already know about those sources, such as https://www.youtube.com/watch?v=hVimVzgtD6w#t=351 ).
The setup would be
1) Create a scatter plot with
a) the time => go to animed dimension to set
b) the bubble your entities you want to compare
c) the color of the bubble (example : the tool/software)
Expressions would be the exact coordinates within the "Gartner Universe"
3) Size of the bubble
However, I'm wondering if not a line chart would better transport the message. Animated chart may look spectacular (in the end humans are mainly visual centered beasts) but the animation might cover the message you want to bring forward. Thus you might set this up using the absolut values of distance to the "Gartner Sun", such as
or you could point out the change over time. For every Tool/Software you would define the original starting point (using again the absolut value of distance to the "Gartner Sun") as 100% and then show as a percentage to this starting point the following positions. The graph would look like
One interesting avenue to explore would be the question, if you should use just the distance to the "Gartner Sun" or if you should use the area that is spanned by the position of the "Gartner Sun" and the position of the Tool/Software (calculated by the formula: area = X * Y). Then there's also the question, where to set the position of the "Gartner Sun". Currently you have set this in far left and below corner of the Quadrants, but you might consider changing this to the middle of the Quadrant map. Using X*Y and keeping the +/- signs of the X and Y might show quite good differenziation of the data.