Address model.water reporting accounting inconsistencies

[Wegatriespython]

The current model.water.report file on the main branch has several issues which need to be addressed :

1. Units are still expressed in Km^3 and need to be swapped to MCM
2. Hardcoded hydro withdrawal multipliers without clear connection to technology specific performance figures. Matching based on string slicing, brittle.
3. Urban Rural Infrastructure lists mixed up :
   • Persistent mixup where rural values are tallied under urban.
4. • Desalination Accounting omission :
   • Membrane electricity input excluded
5. Water treatment electricity consumption mix-up :
   • total water treatment missing urban component
   • rural treatment energy excludes urban data
   • rural transfer energy includes urban efficiency data instead of rural.
6. • Double counting :
   • duplicate rural_mwdem_unconnected, missing industry_mwdem_unconnected_eff
7. Water price hardcoded unit conversion :
   • Division by 1000 under assumption of km3 unit.
8. Sanitation calculation error :
   • pop_drink is used erroneously instead of pop_sani
9. Pop_tot calculation based on R11 will break with R12
10. • Scenario timeseries have population data till only 2015, drinking water and sanitation access assume its available for full horizon.
11. Cooling incorporation.

[Wegatriespython]

Leaving these for a follow-up :
Beyond these specific issues certain quality of life things can be added :

• Strict water balance assertions.
• Accounting completeness, instead of cherry-picking non-zero values for measures (error-prone and difficult to validate)
• Clear separation of model endogenous and post-processing values, current reporting lumps the two together, given a false impression of model endogeneity.
• Legacy reporting decoupling.

[Wegatriespython]

• Allow for Legacy reporting to use cooling reporting in a standalone manner.