"We know that tests on some deconstructed samples have had elevated levels, but we believe that the results of deconstructed tests have little or no bearing on the safety of the finished product," Lynch said.
ARB has an SOP (Standard Operating Procedure) where you take the surface layer off to test the formaldehyde level in the core. So when you have five labs follow the SOP and remove the surface layer - and all show levels of formaldehyde higher than the CARB 2 standard, you have a problem. And when it's 1,000% higher, you have a big problem!
When you have five labs all doing it the same way and getting the same results, it isn't about the test method. The company either didn't understand the SOP, or did understand it and did a work-around. They're either stupid or they're lying - which is it?
The California Environmental Protection Agency: Standard Operating Procedure for Finished Good Test Specimen Preparation Prior to Analysis of Formaldehyde Emissions from Composite Wood Products document (source 1) that we sourced from the official CA site on testing methods (source 3) tells us specifically that deconstruction is a key part of the process.