The Dalitz plot distribution for D⁰ → π⁺π⁻π⁰ produced by D_DALITZ does not match the distribution shown in [[ https://arxiv.org/abs/hep-ex/0703037 | PRL 99 251801 (2007) / arXiv:0703037 ]] (the BaBar paper from which the amplitudes and phases of the resonances are taken). This seems to be due to the fact that the lineshape functions used by BaBar for the Daltiz plot fit are defined differently with respect to the ones used in EvtGen. As they only seem to differ by a scale factor, the amplitude coefficients need to be adjusted by such factors.
All details and correct values of the amplitudes in these slides {F2017503}.