IMO, the world will follow a decline curve similar to the Lower 48 (where we have also tried all kinds of secondary and tertiary recovery techniques and horizontal drilling).  

However, I predict that some countries, such as Russia and Mexico, are poised for very steep declines, and I think that overall net export capacity will be a severe problem starting this year.  

I cannot speak for the world's depletion rates, but the UKCS North Sea is an ideal case for observation.
1) The data is exceptional. 2) Start, peak and tail are a time and geographically bound event 3) Most oil is a light sweet crude and flows well 4) Most of the good oil reservoirs are clastics (sand grains with pore spaces, relatively little cementation). This helps flow rates. (Most reservoirs in the Middle East are Carbonates with very different characteristics for permeability and porosity).
Almost all are offshore and took significantly large amounts of capital to develope in the form of large, hostile weather proofed structures.

Once discovered and developed, production flows were ramped up. The reasons were Financial and Political.
The Oil companies required a prompt return on capital expenditure (big platforms cost money). The Thatcher Government required as much money as they could get to help offset the costs of slash and burn of the older , rust-belt type industries, recession, loss of tax revenue, increased social security burden etc.

Each oilfield was essentially well bounded, understood and recoverable reserves relatively easy to calculate. If you look at production curves for most individual UKCS fields, then you see three phases: Intitial ramping up of production, followed by a brief plateau phase, followed by a decline phase with a slope which is gentler going down than the initial ramping phase going up.

It is likey that overall production (area under the curve) does not increase by much whichever way you produce the oil. Though this is highly debatable since high ramping can inflict damage on resevoirs and producing wells. It is fair to say that operators did take this into account and optimised flow rates to avoid killing the goose.

Oil extracted can be either slow and steady, or fast and furious. The extractable amount in place does not budge much. What happens next is infill drilling to hit sweet spots missed in the intial development. Horizontal / Extended reach / Geosteering drilling all helps. Also, you can occassionally identify stranded , isolated geological traps that contain oil and drill to them from existing platforms (common). Also, maybe you can drill deeper and pick up other oil bearing strata that was missed in the initial development phase (rarer).

However, what is clear is that no field ever went through a
'renaissance' where the initial maximum peak was either matched or equalled. The classic Hubbert curve is exhibited in almost all fields. The curve may be skewed, but the eventual outcome is always the same.

The Chessboard scenario of discovery sequence is also fairly evident in the UKCS. Initially each King or Queeen is developed, and the knights and pawns in smaller , stranded traps become attractive and are developed.

Decline rates on the downward slope after a field peaks in the UKCS has caught a lot of people on the hop. Not least the UK Gov. UKCS appears to have peaked circa 1999. Actuality occurred at least 5 years before theory.
Some of the individual field production graphs are very startling and can vary from 5% per annum YOY to 13% per annum YOY.

Go to Matt Simmons Website and look through his power point presentations. He has quite a few production curves for The UKCS. They are quite a good graphic representation of just how fast a 'major' field can peak and decline.

Carbonate reservoirs do behave differently to clastic reservoirs. The bulk of the middle eastern oil is trapped in Carbonates. One of the main reservoirs in the USA is/was the Austin Chalk. This was very significant in the history of the lower 48 (and indeed the history of the world and especially World War Two...) and I am sure 'WestTexas' could illuminate the story of the Austin Chalk better than I.

Yes, depletion can happen very fast: When we started , we were listening to the Sex Pistols and Souxsi and the Banshees. Morris Marinas were common (ugh). My first car was a Richthoven Red Truimph Dolomite 1500 cc with twin overhead camms...