Re: What is "reasonable" anyway?

You are right that the term, "reasonable" is very fuzzy and must be clarified.  Below I give a concrete example of how the term has been manipulated and a suggestion for clarifying it.

We know from the litigation that arose (and continues to arise) out of the ITU-T's V.34 and V.90 modem standards, that patent holders will attempt to define "reasonable" by arguing that they should be able to recoup their technology development expenses.  Their argument is straight forward enough; the patent system was created to encourage individuals and companies to incur costly development, and so, any definition of a "reasonable" patent royalty must take development cost into account. Although it's a simply argument, it is fundementally flawed in the ITU world, and the W3C should take steps to make sure that such an argument is even more flawed when it comes to W3C standards. 

In the modem world there are 10's of parties that claim patents necessary to implement a technologically relevant modem (this number grows each time a new modem standard is created because of the need to assure backwards compatibility).  If you add all of their royalty demands together you are looking at paying far more in patent royalties than the total bill of materials for the modem itself.  The result is that the modem market is "monopolized" by those companies that have been able to either (i) insert their patented technology into one of the modem standards or (ii) acquire essential patents at a grossly inflated cost from those that have had their technology incorporated into a standard.  This is directly contrary to the stated intent of the ITU's patent policy, which reads, in part:
  ITU-T Recommendations are non-binding international standards. Their objective is to ensure compatibility of international telecommunications on a worldwide basis. To meet this objective, which is in the common interests of all those participating in international telecommunications (network and service providers, suppliers and users) it must be ensured that Recommendations, their applications, use, etc. are accessible to everybody. It follows, therefore, that a commercial (monopolistic) abuse by a holder of a patent embodied fully or partly in a Recommendation must be excluded. To meet this requirement in general is the sole objective of the code of practice. 
http://www.itu.int/ITU-T/dbase/patent/Patent_Policy.html

As clear as the wording in the ITU-T's policy seems to be, in reality, it should be much stronger.  One problem with the ITU's wording is its reference to "commerical (monopolistic) abuse" is vague.  The other problem is that even if an individual patent holder is being "reasonable" the cumulative cost of licensing all of the necessary patents from all of the patent holders often is not.  I suggest the wording to solve both problems:
  No party claiming to hold rights to patents that are essential to implement a W3C standard may charge a royalty for such patents that, when viewed cumulatively with the royalty demands from all other parties claiming to hold rights in patents that are essential to implement the standards, put any party at a material cost disadvantage when competing in the marketplace.  Each signatory of this W3C Patent Policy acknowledges that it is contractually limiting its ability to assert its patent rights by signing this Patent Policy.