"conservativeness"in response to scientific uncertainty and incompleteness: REDD vs. LUC

I recently read a new article from the journal Environmental Research Letters called "Applying the Conservativeness Principle to REDD..." 

The authors argue for a "conservativeness principle" to construct and monitor schemes to Reduce Emissions from Deforestation and Degradation (REDD).  They argue that it's best to make conservative estimates of the emissions reduced by interventions to avoid deforestation because data on the subject are frequently incomplete and/or uncertain.  That way: 

-it's unlikely that the payments will be wasted

-there is incentive for tropically forested countries to prove that the default estimates are conservative

-such programs will then engage tropically forested countries in international climate change governance

I think one can conceptualize REDD emissions reductions as a function of carbon pool*intervention*persistence of intervention.  Therefore, by the conservativeness principle, argue the authors, it is best to conservatively estimate all three of these values.  They note, moreover that, "it is likely that the most typical and important example of incomplete estimates will arise from the lack of reliable data for a carbon pool." As a result, they argue that conservativeness in estimating the size of the carbon pool is the most prudent approach to REDD monitoring.

 It's worth noting, however,that with respect to broader LUC governance this is the opposite of conservative.  With respect to numerous other LUC issues (like say calculating the biofuels ILUC adder) overestimating the carbon stocks more closely falls under the "conservativeness principle".  An overestimate of forest carbon stocks increases the cost of deforestation and puts the burden of proof on industry to advance the carbon accounting science.  Such carbon footprint overestimates are used as default values in the Renewable Transportation Fuels Obligation (RTFO).  

I think I like the premise of the conservativeness principle, but maybe it is best applied to measuring and monitoring the influences of the policies themselves, not the environmental systems/stocks/flows on which they are intended to act.  Otherwise, there is a risk of some strange compartmentalization in the regulatory science.  Thoughts?

FYI, the article also features a decent survey of the scientific literature on carbon flux from tropical forest conversion.

http://www.iop.org/EJ/article/1748-9326/3/3/035005/erl8_3_035005.html