Please consider using something other than the logistic curve. Use a rate-based model and you can actually make sense of the numbers.  The logistic curve is not rate-based.
Of course, you are perfectly right. The logistic curve contains some important assumptions which do not correlate very much with reality. I think the base assumption is that the oil producers are trying to maximize production in every moment of time while constantly drilling more wells to offset decline. The reality of course is not that simple; there are lots of mainly economical restrictions on constantly investing in a field. Usually there are huge initial investments leading to a "spike" of the
initial production. Then the producers tend to invest in expanding production depending on market conditions (which are rather unpredictable). When the field reaches its peak, in most cases it is much more reasonable to maintain a sustained decline than to fight the peak - nobody would pour huge money in a declining field, because the investments are not likely to pay off. For example I am sure that if they drill more wells in West Texas they can temporary rise the production to say 1.2 mln.bbd but then it is inevitable that the production will again slowly slide down (following the logistic curve) and the slide will soon become uncontrollable. That's why they preffer to have a long controllable decline than making unfeasable investments. In short the logistic curve is what we can expect from a field if we target maximum production all the while during its lifetime.

I think that in the short term we will see the West Texas pattern world wide - the first several years will be of significant constant decline rates (3-4%?). But after that the shortages will attract huge investments and in the medium turn the production curve will start to follow closely the logistic curve as producers scramble to boost production by all means.

Of course all of this is too speculative for me to bet on... If the producers start chasing their long-term interest they will probably abandon the "maximum production" goal and even allow temporary free fall of production to preserve it for the future. This is what may cause the real trouble and why we need sufficient energy independance now.

but then it is inevitable that the production will again slowly slide down (following the logistic curve) and the slide will soon become uncontrollable.

Why again does it follow the logistic curve?

You continually critize the Hubbert curve, yet as far as I know you still haven't done any evaluation to show that your model can project forward any better.  Your modeling tends to use a lot of parameters, and smells strongly to me of overfitting the past data.  The onus is on you to show that your model is useful predictively.  I for one am weary of all the carping about the logistic model, which, while it's undoubtedly a crude approximation, does actually have a track record of useful prediction in a significant number of basins.  Please put up or shut up.