Khebab - thanks for this.  I guess there are two messages here.  The first is that the logistic decline looks more rapid than the exponential and that my Mearns2 model may prove over-optimistic.  The second that we need to continue to be wary of data.  I think the data you have used will be for crude+condensate+NGL but may exclude the miscellaneous liquids that arrive at pipline terminals.
1. The decline rate of a logistic curve is proportional to the oil reserve fraction minus the cumulative production fraction(see here for details):

(dP/dt)/P=K(1-2Q/URR)=K(Reserve Fraction  - Cum. Prod. Fraction)

Consequently, the decline will be slow near the mid-point and will accelerate toward K when Q tends toward the URR.
2. If you have another dataset, I can give new estimates.

P.S.: have you received my last email from yesterday?