i think th best solution is by looking at the number of transaction per day, those that are "cold storaging" will end their work at the end so if the number of TX per day rise, it must be because new people joined in
No, bro.
As at the time of stress testing and spam attack there were much much greater number of transaction per day, so this is also not a good way to measure Bitcoin popularity.
Nor only this but one person can make more than one transaction per day. So he can buy four things with bitcoin and here we are with four transaction from only one person.. Such thing can be made by numerous people. The result. Popularity increased and multiplied per four, five or ten times.