A 'private tax-payer test' for State aid? ... or how the Commission is not getting it (about the Apple APA case)

Thanks to @Detig's twitter encouragement, I have finally set out to read the recently released 11 June 2014 Decision of the European Commission SA.38373 in the case of alleged Irish aid to Apple due to the treatment of its advanced pricing arrangements (APAs). Generally, this is a case that pushes the boundaries of State aid law as tax sovereignty is concerned and may force some interesting developments. However, in the particulars, its seems that some of the foundations of the Commission's position are rather shaky.
In my view, one of the points where the Commission's logic is particularly flimsy comes when it tries to justify the application of the private operator/investor test in this context, in what should be rebranded as 'private tax-payer' test, by stressing that 'to avoid this type of advantage [ie the allocation of profit to subsidiaries in low tax jurisdictions] it is necessary to ensure that taxable income is determined in line with the taxable income a private operator would declare in a similar situation' (para 9, emphasis added). This just does not make sense and incorrectly focusses on the incentives of the economic operator (tax payer) instead of those of the tax authority (which, in the end, is the one that may have accepted APAs that granted an undue economic advantage to the former).
 As the Commission had itself very clearly indicated (para 8 of the same document), the financial incentives that (multinational) private operators have are exactly in line with Apple's behaviour. Hence, the Commission should have stuck to the simple truth that, from an economic perspective, the only rational behaviour that can be expected from economic (corporate) operators is to try to minimise fiscal pressure and to incentivise their tax directors to do so [Armstrong, Blouin & Larcker, 'The incentives for tax planning' (2012) 53(1) Journal of Accounting and Economics 391-411].
This may not be the socially desirable behaviour, and precisely that is why tax law is there [as, indeed, 'if we were ideally virtuous, there would be no need to study what people should pay in taxes to finance subsidies to the poor, the employment of a police force, and provision of an urban infrastructure, or to find ways of reducing the environmental damage we do'; J Mirrlees, Welfare, Incentives, and Taxation (Oxford, OUP, 2006) iv]. 
If the Commission is of the view that the activity of the (Irish) tax authorities was not in line with rational behaviour, it should not try to find a justification in the behaviour that could be expected from the tax payer, but rather on the rationality of the decision of the tax authorities on the basis of the existing knowledge on optimal taxation--an issue discussed by Mirrlees (131-73) and many others, without having necessarily reached a final conclusion so far [see an interesting discussion of the main insights achieved so far in NG Mankiw, M Weinzierl and D Yagan, 'Optimal Taxation in Theory and Practice', (2009) 23(4) Journal of Economic Perspectives 147-74].
Trying to conflate this insight and to word the criterion for the assessment of Apple's APAs as a 'private tax-payer' test does not make sense and risks damaging the consistency and logic behind the principle of private operator/investor test as a general principle for the assessment of State aid [for discussion, see A Sanchez Graells, 'Bringing the ‘Market Economy Agent’ Principle to Full Power' (2012) 33 European Competition Law Review 35-39].
In my view, this is plain to see in the oddity of the detailed reasoning in which the Commission engages, when it establishes that
When accepting a calculation method of the taxable basis proposed by the taxpayer, the tax authorities should compare that method to the prudent behaviour of a hypothetical market operator, which would require a market conform remuneration of a subsidiary or a branch, which reflect normal conditions of competition. For example, a market operator would not accept that its revenues are based on a method which achieves the lowest possible outcome if the facts and circumstances of the case could justify the use of other, more appropriate methods (SA.38373, para 56, emphasis added).
Quite honestly, it is very difficult to understand what the Commission exactly means by this--and this is the more worrying because '[i]t is in the light of these general observations that the Commission will examine whether the contested rulings comply with the arm’s length principle' (para 57). If what the Commission indicates is that for the purposes of taxation, a rational/prudent economic operator would not accept a method that results in the lowest possible tax base, this just does not make sense. Differently, if what the Commission means is that for purposes other than taxation (which would those be?) the rational/prudent economic operator would equally oppose that method, then much more detailed explanation of why and how that is the case would be needed.
Worse of all, the Commission has a strong cases on the facts. The Irish tax authorities entered into negotiations with Apple and allowed the company to deviate very significantly from the applicable (general) tax rules. Moreover, despite the very significant development of international standards on transfer pricing, a 1991 ruling was used until 2007 with no revision. This sweet deal for Apple was clearly linked to an objective of keeping (regional) employment and ensruring some tax income. These may be rational (?/justifiable?) political decisions, but they do not meet any acceptable standard of objectivity, professionalism and transparency and, consequently run against the basic requirements of good (tax) administration. And, what is more important, clearly point towards a selectivity in the application of the tax system that makes the whole deal fall foul of the prohibition in Art 107(1) TFEU [the important legal point is, indeed, made at para 70 of the Decision].
In view of all this, one cannot but wonder why would the Commission base its case on such unfocussed and difficult to share (to put it mildly) points of departure. One possible option, of course, is the rebalancing of powers in tax matters derived from the Treaty of Lisbon and the very limited space for action in the front of direct taxation that is not supported unanimously by the 28 Member States (see art 115 TFEU) [for discussion, see TA Kaye, 'Direct taxation in the European Union: from Maastricht to Lisbon' (2012) 35(5) Fordham International Law Journal].
Another possible option is that the Commission is trying to deflect the bad publicity from the Member State concerned (Ireland) towards the multinational (Apple), hoping to find less resistance (or to trigger support) at Member State level. There can be a myriad other reasons, of course. But none of them seems to justify risking a case (and a principle of enforcement of State aid law) in an attempt to get the prohibition decision through.