`report.gms`: Fix `hours` for pre-final-year stress periods

[patrickbrown4]

The handling of hours(allh) in report.gms is currently a bit off for stress periods:

• hours(allh) includes both representative and stress periods and is redefined in 2_temporal_params.gms (and upstream in hourly_writetimeseries.py) for each model year and ReEDS/PRAS iteration
  • So after the last year/iteration, it only contains the stress hours used in the last year/iteration (typically 2050)
• But in report.gms, there are a few places where we multiply it by stress periods (h_stress_t(allh,t)) across all years:
  • tran_util_ann_stress
  • net_import_ann_stress

I think the best approach would probably be to add a new reporting parameter like h_stress_t(allh,t) (something like hours_t(allh,t) and then use that in report.gms.

[patrickbrown4]

Addressed by #151