When you look at a production rate versus time graph, if you integrate the area under the curve, you get URR, or Deffeyes' Qt.

The question is, what is the area under the curve, especially when the graph shows primarily rising production with time?  

There are a lot of engineering terms--proven, proven undeveloped, probable, possible, etc.  The USGS uses some very optimistic methods to get their reserve estimates, and I recall that someone at Saudi Aramco suggested at one time that they were using USGS estimates.

However, I prefer the Hubbert Linearization (HL) method to estimate Qt for large producing regions and for the world.

Using the HL method, if we have a enough production history, we have been able to demonstrate pretty accurate results.   I would especially point to the Lower 48 case history, where Khebab took the production data through 1970, to generate a post 1970 predicted production profile.  The post-1970 cumulative Lower 48 production was 99% of what the HL model predicted.

We have consumed about 1,000 Gb of crude + condensate, and Deffeyes estimates that the world has about 1,000 Gb of conventional crude + condensate left.   IMO, the only areas left that could really change this estimate would be the north and south polar regions.  

In any case, as predicted by Deffeyes, world oil production has been falling since December, and as Khebab and I predicted, production by the top oil exporters has been falling faster than world oil production is falling.  

I really can't think of a time when Yergin and/or Lynch have been right.

I can point to multiple examples of the HL method being correct, especially for large producing regions.

Take your pick.