That's one way to look at it, though I do feel this case is somewhat unique.
The 10xx series was very good. It offered upgrades at every price point (disregarding the miner wars, ugh). Someone who had bought in at a particular price point a year or two previously had new choices at the same price point worth considering.
20XX is, so far, mostly a wash at the same price points. What this means is that Nvidia is by default in competition with its own past. A buyer of a 1080, 1080ti, 1060, 1050ti, etc have no choices in new cards that offer a compelling reason to do so. Given how lengthy the 10xx shelf life was, further enhanced by great AIB models for the majority of that time, this is a massive swath of their consumer base that has no reason to look at RTX.
Ditto owners of RX470/480/570/580/590/Vega56/Vega64. Unless they are jumping up in budget considerably, there is nothing new on offer that gives them more than what they already have.
These factors are going to hurt Nvidia sales until something changes. Extreme .1% type gamers and all new builds with beefy budgets are not enough to meet their expectations in regards to sales. RTX features were apparently supposed to be the lynchpin for getting those wider-market buyers on board, but so far that is a big zero, and the prospects of that changing seem extremely bleak.
It's a weird situation, and Nvidia doesn't really have many good options other than plan for a good 7nm jump with less die space pushed for tensor as soon as reasonably possible, which is likely to be at least spring 2020 if not later.