There is the following rule of thumb in Malhi et al. ("The productivity, metabolism and carbon cycle of two lowland tropical forest plots in south-western Amazonia, Peru", Plant Ecol & Diversity). I imagine the Galbraith paper is referring to this '39% rule':
"We also estimated that there is an additional 39% fine root NPP beneath our 30 cm ingrowth core, by extrapolating to 1 m depth assuming an exponential decay of root biomass and productivity with depth"
this rule is equivalent to assuming 61% of all roots occur in the top 30 cm of the soil profile and may also be estimated from an equation used in the CLM model (as pointed out by Chris Doughty: see footnote p.58 of the Manual v3 and the same is encoded in rootbiomassf_s.m in the equation for dzz).
I asked Dan Metcalfe about this 39% rule in 2012 and he said it came from data collected by Rosie Fisher at Caxiuana. However I emailed Rosie F at that time and she said that this didn't come from her data so there has always been a bit of confusion over where the 39% rule originally came from.
After looking into this more deeply I found that Jackson et al. (1996) put the percentage of roots above and below 30 cm as 69%-31% rather than 61%-39% and that this is backed up by studies across tropical humid forests. That is why Jackson's estimate appears in the Table on p.58 of the Manual v3 and I recommend this to be used over the 39% rule.
Therefore, in answer to your question: that 39% rule is used by several references, but I suggest rather to use the 31% rule of Jackson (which is in the GEM manual section 2.3).
Hope this helps.