Of course, prior to Colorado and Washington's cannabis legalization, there had been a widespread belief that before-tax prices would plummet by at least half, and perhaps even 80% or more, upon legalization or very shortly after. And yet, that did not happen--yet. If anything, prices went up at first due to higher demand in the short term, and with or without taxes the prices three years later still have not dropped nearly as much as predicted compared with the status quo ante, at least not at the retail level.
And while the price elasticity of demand for cannabis remains controversial to this day, it is generally thought to be relatively inelastic overall, though estimates vary dramatically. Note that all estimates were calculated during the time of prohibition and do not always adjust for potency. There may very well be a relatively "elastic zone" above a certain price point, such as the average price during the 1980s and early 1990s when usage rates dropped, but any further decreases from current levels are squarely in the "inelastic zone". In fact, even before legalization, the typical cost per buzz-hour was already cheaper than beer.
As for heavy use, the definition of that term varies, but for argument's sake let's use Mark Kleiman's "daily or near-daily" (DND) use definition of 20+ times per month. He notes how the rate of past-month use among adults is roughly the same as it was in the early 1990s, but within that group the proportion of DND users has apparently increased fourfold by 2014. And yes, heavy users would be the most sensitive to price as a rule, since they spend a more significant chunk of their incomes on weed than non-heavy users. But this increase occured almost entirely during prohibition, and now it seems like the market for heavy use is quite saturated even with legalization and falling prices. So heavy use is unlikely to increase much more (if at all) going forward, regardless of price or legal status.
Thus, an 80-90% price crash at some point in the not-too-distant future is really not too farfetched. If that happens, then the average potency would also likely drop from its currently high levels. What would be the point of using energy-intensive indoor growing methods to force the THC percentage levels well into the double-digits and even into the twenties, when decent bud can be produced much more cheaply and efficently via outdoor growing or in literal greenhouses? After all, the whole point of high-potency weed is more "bang for the buck", as well as making it easier to conceal. At the same time, bunk weed would be, well, weeded out. And specialty products such as edibles and vape products will likely increase in production, as the cost of producing them will go down due to cheaper crude cannabis prices.