> OK, what difference in general one would expect
> for the transition energy calculated with and without
> the recommended EDSCHFT, which is 0.02. In my case I get the
> difference of about 0.1 eV. Is that acceptable/normal in general?
In most cases QDPT calculations are plagued by intruders so that
the difference up to several Hartrees could be expected while
comparing computations with and without ISA shift.
The difference of 0.1 eV is in general acceptable and likely
indicates that there are no significant intruder states-related
problems in your calculations.