Thanks, Emmanuel. I know this area very well. Each reaction rate low is valid in a certain range of concentrations. Surely that at [O2] = 0,the rate is = 0. With an increase of [O2]. the rate increases and then reaches a plateau. I gave a ref on old paper in order to show that a zero order with respect to O2 is known from 50s. The parameter (kp)^2/(2xkt) ( kp is for the reaction RO2 + RH, kt is for RO2 + RO2 --> keton(aldehyde) + alcohol + O2.) is commonly called the "oxidizability" of hydrocarbons.
the concept of the reaction rate order has been introduced a very long time ago. Later on, "the reaction rate law" has been commonly used. However, in numerous cases and even for a rather simple reaction mechanism you can't derive the formula for the reaction rate law. In some pretty rare cases under certain experimental conditions the reaction rate law simplifies to a classical equation: rate = k([A]^m)([B]^n) and you can determine the apparent reaction rate orders. In modern chemical kinetics the main goal is to write the reaction mechanism with reaction rate constants assigned to each elementary step. The fitting (simulation) of experimental curves can be performed using digital integration. The problem is considered to be solved, if experimental data/curves are consistent with calculated. From this point of view, the question on the apparent reaction rate order is not important. Elementary chemical reactions could be only uni- or bimolecular (you can't call them the first or the second orders). In this context the term "true zero order" does not make any sense.