In my understanding "the effective pH in water" is nothing else but the Hammett acidity function, which is extension of the classical $\text{pH}$ quantity to non-aqueous solutions. The blue ranges are then available Hammett function values for the given protic solvent, similarly as there is range of available pH values for water solutions.
The respective dissolved conjugate acid/base pairs can have in other then water protic solvents such activity ratios that would hypothetically belong to the "effective pH in water", if such ratio should be achieved in water solution.
E.g. $\ce{HNO3}$ has reportedly $\mathrm{p}K_\mathrm{a} = -1.5$. If it was dissolved in glacial acetic acid(preferred IUPAC name over the IUPAC systematic name ethanoic acid) and the solution was adjusted in a way $\ce{HNO3}$ : $\ce{NO3-}$ ratio was 1:1, then the solution in acetic acid would have the "effective pH in water" equal to -1.5.
IF the conjugate acid/pair activity ratio belonged to the effective value out of the blue range, it would be the acid is stronger than protonated solvent and it would protonate the solvent. Or, the base is stronger base than the deprotonated solvent and would deprotonate it:
\begin{align}
\ce{HA(solv) + Solv(l) &<=> A-(solv) + HSolv+(solv)}\\
\ce{B(solv) + HSolv(l) &<=> BH+(solv) + Solv-(solv)}
\end{align}
The left edge of blue ranges maps itself to the case $\ce{HA(solv)}$ and $\ce{HSolv+(solv)}$ are equally strong acids.
The right edge of blue ranges maps itself to the case $\ce{B(solv)}$ and $\ce{Solv-(solv)}$ are equally strong bases.
The edge values can be considered as $\mathrm{p}K_\mathrm{a}$ of protonated(left) and not protonated(right) solvent.