But:
a) in that case I don't see where the 800kbpd of Manifa in 2012 went, since there is only one small blip, then flat, and Khurais should be still increasing.
b) you have a kick upward in production this year, being attributed to existing fields. You also have news stories saying Khursaniyah is in full production. Swap the 'old' for 'new' in light of that and your future production on old fields extrapolation gets pulled down significantly across the future predictions.

Also the megaproject page has Khurais down against early 2010. Even if believe the Saudi date of mid 2009, your own satellite work suggests more of a 'big bang' startup does it not?

I agree with the aim, but I'd be included to assume the behaviour of the 'old fields' should be a punctuated, smooth decline - with certain fields maybe being shut down. If you are not getting that, I'd question if the data for the other curves passes the sniff test - particularly in light of stores as a buffer that's not included.