From the Oil Cast interview we have:
One has to realize that in order to replace the capacity of Cantarell with these new fields, (none of which exceeds 200 million barrels) we would need some 20 new fields. Right now we have only discovered three.

Suggesting that new fields are hardly going scratch the surface of the Cantarell decline and would do well just to substitute for non-Cantarell decline. We also have this statement:
The fifth module of the nitrogen plant, located in Atasta, off the Campeche shore, will be used in the pressure maintenance of the field. It is probable that Ku will reach some 800,000 barrels at its maximum. With this production, the Cantarell decline will be mitigated a little bit. But the problem with Ku is its extra heavy oil.

I don’t know what Ku is producing today but it looks like it can increase by a few hundred thousand barrels - but it’s extra heavy. Similar to the oil the Saudi’s can’t give away.

I think the 850kb/d is just a little bit optimistic, dropping it to 600kb/d would just about bring it in line with my graph.

The following is the data on Mexico from the Koppelaar report. It appears that the decline will be attenuated by new fields, and Type III decline will be about 3 to 6% per year in the near term.

Mexico production forecast
2004: 3789
1st qtr 2005: 3825
End 2005: 3663
End 2006: 3428
End 2007: 3316
End 2008: 3324
End 2009: 3218

On Mexico
“In 2004 oil production was 3.38 mb/d and NGL production was 442.000
b/d, giving a total liquids production of 3.83 mb/d. An annual decline
rate of 10% over 388.000 b/d of 2004 oil production has been added. An
annual decline rate of 12% over 2.14 mb/d of 2004 oil production
(Cantarell) has been added. An annual decline rate of 3% over 548.000
b/d of 2004 oil production has been added. NGL production was assumed
to remain stable. On top of this a total of 850.000 mb/d from 2
projects has been added.

Specific information: “According to then Exploration and Production
Director Ramirez Corzo, Cantarell’s production should remain stable
until 2006, but would decline by 14% per year after that. However,
Ramirez Corzo recently stated on November 2, 2004 that “Our best
estimate is that Cantarell will start to decline toward the middle of
next year [2005],” raising the possibility that Cantarell’s decline
could come sooner than originally had been thought.”
http://www.peakoil.nl/images/ponlreport.pdf