There seems to be some discussion on using solubility as an indicator of roast development. See Ben Kaminisky's talk at NRF13, read http://vstapps.com/blog-2/extractmojo/underdevelopment/.

My goal would be to target a solubility for all roasts that one grinder setting could be trusted to extract all my coffees equally.

Now, how would one construct a solubility test protocol specific to coffee?

A few concerns come to mind, mostly with regards to tight consistency:
temperature consistency... for instance, cupping bowl thermal mass
CO2 emission interfering with extraction
when to draw a sample if from an immersion brew
how to brew consistently if a percolation brew

My first stab is this: My grinder is set to 800microns average particle size; a widely accepted target for brewing. I'm preheating cupping bowls with 150F water from my sinks, brewing 10.2g / 150g, dunking the bloom at 4min, then at 10min pouring the bowls through a dry v60 filter to destratify / filter and then test for TDS.

I have some concern that the filtering of the brew may incur some amount of additional extraction that is not consistent bowl to bowl. I also wonder how the initial pour turbulence affects extraction. I'm sure there are other variables I'm not taking into account. Hence this post.

FWIW, using my protocol gave some results that did not correlate to development. One poorly developed cupping roast (by taste, color, and time), two well developed mid 1st crack roasts, and one edge of 2nd crack roast.

1.39%, 1.48%, 1.45%, 1.35% accordingly.

Either my protocol is not good, or the correlation not true.

All the same coffee?

I am using a syringe at the end instead of pouring into paper filter.
I think it will avoid more agitation and further extraction on the fines in the paper filter.