We are left with a purportedly iron-cast way of thinking that says that reasoned thought about the real world must not be values-influenced, and that thought about values must not be influenced by observations. The idea that this separation is workable is essentially stated without proof. No alternatives are presented or considered.
Realistically, in the case of sustainability issues, a wide variety of values-based and science based contingencies interact. The world does not guarantee us the clear simplification implied by Pielke's taxonomy. What Pielke's taxonomy does offer is a way to express outrage at anyone trying to grapple with the full specturm of challenges.
Just because "the honest broker" is a nice turn of phrase (and it is) doesn't mean that the book of that title is authoritative. In fact, it is just a muddle; more useful to prevent progress than to support it.
To be fair to Roger, in his meanderings he happens upon a quote that seems to me to make a great deal of sense, from Harvey Brooks, "Evolution of US science policy" in Smith and Barfield, eds. "Technology, R&D, and the economy", Brookings Institute, 1995:
In the process of using science for social purposes is thought of as one of optimally matching scientific opportunity with social need, then the total evaluation process must embody both aspects in an appropriate mix. Experts are generally best qualified to assess the opportunity for scientific progress, while broadly representative laymen, in close consultation with experts, may be best quallfied to assess societal need. The optimal balance between opportunity and need can only be arrived at through a highly interactive mutual education process involving both dimensions.Well, though Roger doesn't note it, that's obviously about funding science, not using science, obviously, but most of it applies. The "highly interactive mutual education process" is key. There is no formula for good policy. Bad policy is dramatically easier to achieve than good policy both in enactment and in practice. Trying to bottle the conversation, to constrain people to specific roles, is an extra constraint in a problem that is already, apparently, overconstrained. It just makes achieving good policy all the harder.