Here are some initial forecasts for CMBS4 non-Gaussianity error bars on local and equilateral non-Gaussianity. Orthogonal to follow, as well as polarization constraints.
I have plotted both the signal as a function of and . Equilateral triangles have very little sensitivity to the minimum multipole, but local non-Gaussianities, which rely on one mode being squeezed do. I expect that not having the lowest modes, increases the error bars by 50 %. The numbers below are comparable to the full sky numbers in 0701921v1, Fig. 2 for a cosmic variance limited experiment (temperature only). The normalization I used is from 0405356v1, Eq. (6). I used a flat sky approximation with .