Spring QM lab: full run-through in ECCE with Gaussian 16 and with ORCA #564

Run every step of the spring computational QM lab inside ECCE, once with Gaussian 16 and once with ORCA, on Kebnekaise, and record what fails or needs a workaround. Parity target: what the lab does today in GaussView (#410).

Exercise 1 (Diels-Alder): s-trans butadiene Opt+Freq; relaxed dihedral scan 36 x 10°, ΔE vs angle plot and .csv; frequencies at the scan maximum (New calculation from this point); s-cis Opt+Freq with NBO/NPA charges; ethene Opt+Freq with charges; reactant complex and product with the same atom numbering; TS by QST3 (G16: three structures; ORCA: NEB-TS or OptTS from a guess, #505); one imaginary frequency; IRC both directions, 20 points, path plot; thermochemistry (H, S, G) for the reaction profile; HOMO/LUMO pictures.
Exercise 2 (hydrogen bond, one partner >= 10 atoms): complex and monomers Opt+Freq, NBO/NPA charges, dipole moments, H-bond distance and angle, total density mapped with the ESP.
Exercise 3 (Raman): cocaine and benzocaine Opt+Freq with Raman; both spectra compared in one plot, differing peaks picked (#562).

Points not yet confirmed: QST3 with the guess structure; Pop=Full NBO output shown as charges (Pop=NPA is parsed); ESP-mapped density for G16 and ORCA; ORCA Raman for a 43-atom molecule within the lab's time.

Each failure becomes its own issue. No course or institution names in the repo.