Expand PyWake submodel support with case-insensitive lookup and tests

[bjarketol]

Summary

WIFA's PyWake adapter (wifa/pywake_api.py) currently supports only a small subset of the models available in PyWake. This expands all 6 submodel categories to cover the practical PyWake models, adds case-insensitive name matching, and adds parametrized unit tests.

Changes

New _normalize_name() helper

Case-insensitive model name matching — strips whitespace, lowercases, removes - and _. Handles windIO naming inconsistencies (e.g. TurbOPark vs TurboPark).

Expanded submodels

Category	Before	After	New models
Deficit	5	11	Niayifar2016, Zong2020, Carbajofuertes2018, TurboNOJ, GCL, SuperGaussian2023
Deflection	2	3	GCLHill
Turbulence	4	6	GCL, IEC-TI-2019
Superposition	2	5	Max, Weighted, Cumulative
Rotor averaging	2	6	EqGrid, GQGrid, PolarGrid, CGI
Blockage	3	9	SelfSimilarityDeficit, RankineHalfBody, Rathmann, VortexCylinder, VortexDipole, HybridInduction

Other improvements

• Removed 9 debug print() statements, replaced default warnings with warnings.warn()
• Bastankhah2016 raises NotImplementedError with helpful message pointing to foxes or alternatives
• Product superposition raises NotImplementedError (not available in PyWake)

Tests

• New tests/test_pywake_submodels.py: 104 parametrized unit tests covering all configure functions
• All 10 existing integration tests pass (no regressions)