From an investigation of Meqtrees pointing offset solutions with @o-smirnov. Pinging @bennahugo
The pointing offsets recovered by calico-wsrt-tens's solvable E-jones module seem to be scaled by a factor of 0.27. See the plot of input offset (from a WSRT simulation) VS recovered offset below
The pointing offsets recovered by calico-wsrt-tens's solvable E-jones module seem to be scaled by a factor of 0.27. See the plot of input offset (from a WSRT simulation) VS recovered offset below