If it was complete randomness, then you would see completely divergent paths between the two prospects. The fact that they had the same type of growth in OVR every single season points more towards the direction as things being pre-determined, and with a possibility of statistical based growth.
One prospect having an advantage in statistics doesn't completely remove the possibility that they affect player growth. It could simply be that NHL statistics are taken into account. There is also the possibility that it's not based on absolute stats, but it could be looking at something like production per minute of ice-time, or even a combination of both. It also depends on what their starting attributes were. If both players had good to high offensive attributes, it could be that it's their improvements in defensive abilities that is driving their overall growth, in which case maybe you're looking at the wrong statistics.
The only problem right now is that there appears to be no randomness to begin with. That's why it is better to get Galchenyuk rather than Yakupov. Or even better than that, go after Gaunce, Bartschi or Huberdeau. Consistently across the board, people are seeing the exact same development from recognizable names. Mind you, further on in GM mode where the draft picks are randomly generated, it might be more difficult to predict, but at least in the first few seasons, there is nothing random at all.