The new peak for October (all liquids) is based on the IEA data for which the front month is November, see Stuart post for an in depth discussion:

Does the latest IEA number matter?

The problem with the IEA is that they give only limited public data (only the three last months) and it's difficult to track revisions whereas the EIA can revise its monthly estimates up to three years.

The EIA Short Term Energy outlook (STEO) has also preliminary estimates for October and November:

October: 85.96 mbpd
November: 85.61 mbpd

Trailing v Forward

The world's cereal stocks to use ratio
is even farther behind the curve.

http://www.investmenttools.com/futures/soy/chicago_wheat_futures_market.htm