[QDP] IQP encoding optimization roadmap

[aloha1357]

This issue tracks several IQP encoding optimization experiments.

Potential work items:

• Phase kernel optimization
• Shared-memory FWT path
• Batch throughput optimization
• Kronecker decomposition based FWT
• Tensor Core acceleration
• Implicit Hadamard engine
  The goal is to submit these improvements as independent PRs for easier review.

Related:
#1227