Should FRAND Royalties Be Returned to SEP Licensees if the SEP Is Invalidated?

Stephan Altmeyer and Christopher Weber have published an article titled Rückzahlung von Linzenzgebühren bei rückwirkender Vernichtung eines SEP? ("Reimbursement of Royalties in Response to a Subsequent Invalidation of an SEP?" in the December 2017 issue of GRUR, pp. 1182-88.  Here is the abstract (my translation):

The case law has long held that paid-out royalties are not to be returned, if the licensing IP right is subsequently invalidated.  On the other hand, it has never been decided whether this principle also applies to so-called standard-essential patents.  This essay presents the development of the case law from the Reichsgericht up to the most recent CJEU decisions, and identifies the fundamental legal and business differences between this case law and license contracts concerning SEPs.

The authors trace the general rule, that licensees are not entitled to recover back the fees they have paid to license a patent that is later declared invalid, back to 19th century case law, and then forward to the CJEU's more recent decisions such as Genentech finding no violation of EU competition law from a contractual provision requiring the payment of royalties even after a patent is found to be invalid or not infringed.  One of the rationales of the case law is that, even if a patent is eventually invalidated, up until that point the licensee enjoys a market advantage vis-à-vis non-licensees.  But this rationale, the authors argue, doesn't apply in the case of FRAND-committed SEPs, which can be used by any and everyone subject to the payment of a FRAND royalty.  The authors also argue that, under the present system, the incentive on the part of SEP licensees to challenge potentially invalid patents is deficient--due to, among other things, the cost of challenge, and though the authors don't mention it directly, the fact that the beneficiaries of a successful challenge include the challenger's competitors as well as the competitor itself--such that, as a result, the price of goods incorporating such technologies may be artificially high. 

Overall, it's an interesting argument, though I wonder if the authors' proposal were to be adopted if the upfront rate demanded by licensors would simply increase to offset the risk; if so, then without some regulation of that upfront rate, licensees might not be any better off, over time and in the aggregate, than they currently are.   For previous discussion on this blog, see here.
 
One further note--the general rule in France appears to be the same as in Germany, namely that the licensee doesn't get the royalties back even if the patent is invalidated.  See Alain B. v. Thiérart SARL, PIBD No. 1068, III, 202, 203-04, Trib. de la grande instance de Paris, Jan. 12, 2017 (citing SA New Holand France v. SA Greenland France, Cour de cassation, Jan. 28, 2003); see also Jacques Raynard, Anéantissement du brevet et sort des redevances de license:  efficacité de la clause prévoyant la non-restitution des redevances acquitées, Propriété Industrielle, July-Aug. 2017, pp. 27-30 (critiquing the rule, in the absence of a contractual provision permitting the licensor to keep the royalties).