So if Crude is going down and Liquids going up, what's making up the the increase? NGL? Or are agro-fuels starting to make a difference?

My assumption is that biofuels are making a difference but not in the way you might expect.

The production figures include crude oil, lease condensate, natural gas plant liquids, biofuels and refinery processing gains.

However at least in the USA, ethanol production is not counted until it is blended with finished gasoline. In my view, this means that during mid year driving months including October, the total liquids production (or "oil supply") includes stock drawdowns of ethanol instead of ethanol production. As ethanol production is becoming more significant in the world, the total liquids production figures will tend to be overstated during mid year.

For the USA, please see footnote 2 of
http://www.eia.doe.gov/emeu/ipsr/t14.xls
"Beginning in 1993, (total liquids - my words) includes fuel ethanol blended into finished motor gasoline and oxygenate production from merchant MTBE plants"

Further from the EIA glossary under Motor gasoline (finished)
http://www.eia.doe.gov/glossary/glossary_m.htm#motor_gas
"Volumetric data on blending components, such as oxygenates, are not counted in data on finished motor gasoline until the blending components are blended into the gasoline. Note: E85 is included only in volumetric data on finished motor gasoline production and other components of product supplied."

In other words, in the USA, according to the EIA, ethanol is only counted when it is blended not when it is produced. Hopefully, the ethanol that is imported from Brazil is not double counted.

For comparison, note that the OPEC Monthly Oil Market Report November 2007, page 31, stated that "for the month of OCtober ... world oil supply averaged 85.89 mb/d", over 0.5 mb/d lower than the IEA number of 86.5 mb/d.
http://www.opec.org/home/Monthly%20Oil%20Market%20Reports/2007/mr112007.htm

Also, if I recall correctly, the EIA also does not double count (a change made for 2007 if memory serves) in the production of biofuels. E.g., a barrel of oil used directly to produce 1.2 BOE of ethanol is no longer counted as 2.2 total BOE production. Rather, it is counted as 1.2 BOE.

I think they realized that as EtOH plants really took off and they were relying upon fossil fuels and especially liquid fuels, the numbers would be artificially inflated. Seems I recall reading that in either the notes to the IPM or the MER in late 2006.

I'm a bit dubious about the notion that it is double counting. If x barrels of biofuel use y barrels of oil, then that y barrels would show up on the demand side and x barrels on the supply side. If they're equal, there would be no overall change in supply/demand.

There is no real need to modify the figures when looking at the whole picture, otherwise you'd have to do the same for all oil resources.

In term of worldwide production you are not correct. For example, our 2nd largest supplier of oil is Mexico. Their total production is of course added to worldwide production figures and the oil is delivered here and elsewhere. It shows up as supply. The products of that oil may then be used to farm for ethanol production. The resulting ethanol then is also added in its entirety to the worldwide liquids figure. No where is it shown as a net production figure.

But the oil that went into production of the ethanol would then be extra consumption, thus cancelling out the ethanol production figure. So I still see no double counting. Am I missing something?

Consumption prior to 1 barrel of ethanol is X. Now that the ethanol is produced consumption is X+1. That's still the X that would have been consumed, plus the 1 that went into producing the ethanol. But we have 1 barrel of ethanol to add. So supply is now X+1, whilst consumption has leaped to X+1 also. No overall change from X consumption and X supply.

Net figures are not needed; it all works itself out in the supply and demand equation.

Look again. Figure 1 through October, figure 2 through August.