Great explanation as always Roland.
I do agree with everything you said, but my point of view (and the one i always try to disseminate in my answers) is that manufacturers should take care with their use of standards. There are a lot of problems with using standards or harmonized or recognized standards, and one the problems i´ve noticed by experience, at least here in Brazil, is that when you put in the head of a manufacturer that he HAS to use some or alot of standards (and i know a lot of certification house \ notified bodies which "sells" to a manufactuer that his only option is to use the standards) it results in the manufacturer having a "narrow view" of the whole regulatory scheme system, and this has a negative impact on the safety and effectiveness of the device (which is generally what reg schemes are about). This happens because, among other things, manufacturers tend to think that only complying with a standard or some standards will make their equipment safe.
Take your biocompatibility example. A lot of manufacturers tend to think that all biological safety can be dealth with using the ISO 10993 series, mainly because they are are known standards and certification \ notified bodies always asks for that. But, although biocompatibily IS a requirement from the MDD (explicited in 7.1 of the essential requirements), the MDD is in fact not asking just for that (it´s obvious when you note that 7.1 says "Particular attention must be paid to: .....some points" - it does not mean that you have to pay attention to ONLY those points). In fact, what the requirements asks is for a evaluation of biological safety, and biocompatibility is only an aspect of biological safety (for example, biocompatibility is concerned with interactions between tissues and materials, with regard for example for irritation - it´s not concerned, if the material is poisonous or not, so there can be poisonous materials with are biocompatible....clearly this is not what the regulatory requirements want!).
But, as everyone talks about biocompatibility standards, ISO 10993, and everyone asks just for biocompatibility studies, manufacturers are clearly overshadowing the "big picture" concern of biological safety (again, this is from my experience in Brazil mainly, but i know that this happens abroad too). This is an example of what i fear when using standards.
So, again, i do agree totally with you, and would indicate that manufacturers should always use standards because they facilitate a lot of things (including assessments from regulatory authorities). But i will also advise that they know what they´re doing, because blindly following the advice of using standards without really knowing why they can use it, or why it´s best to use them (even if they´re not compulsory) is a risk for their business and might put them into trouble sometime.