It would be interesting to run some thought experiments or even more rigorous simulations on how current GPU plotting difficulty may have an impact on mainnet growth trajectory.
Assume rational miners who would only join mining when (its reward) * (token price) > OPex + CAPex over a certain period of time. Lets expand left hand side which is roughly a function of (miner’s space) * (token price) / (network total space). Given this relationship, if network total space grows exceptionally fast, say for example doubles in 2 weeks (GPU plotting) but token price stays relatively the same, it maybe that the miner ROI would crossed threshold and stagger the speed of new player joining the network. Whats worse, a miner may did the math on day 1 and bought the HW but over the span of 2 weeks it may find out that ROI is not sustainable anymore. Of course, there are tons of scenario to play with but by current plotting difficulties it maybe fairly imaginable to have fast growing network total space given what we know about the amount of GPUs are available for mining.