Remove pT thresholds in combiner for DiElectronSoft

Actually this is a follow-up of !2043 (merged) where the DiElectronSoft line was forgotten for an unclear reason. The code change should really be small, the rate is expected to go up a little but not too much. The line was never intended to have any pT cut as the strange decays have electron pts that go down to the limit of reconstruction (~100 MeV) I tested with meps from May 30 2025 and the sequence hlt1_pp_forward_then_matching_and_downstream_with_parkf_1200kHz_M23

Before change:

Hlt1DiElectronSoft: 1/210000, ( 0.14 +/- 0.14) kHz

With using the nopt combiner:

Hlt1DiElectronSoft: 4/210000, ( 0.57 +/- 0.29) kHz

With loosening a few cuts moderately to gain efficiency:

Hlt1DiElectronSoft: 8/180000, ( 1.33 +/- 0.47) kHz

The cuts were tuned by hand following the studies by Pongsakorn last yearCERN_Summer_student_report_Pongsakorn_Sriling.pdf