[Epic] 4.A GP Temporal and State-Space Methods

## Theme
Bring temporal and state-space GP structure into the package through kernel-to-SSM conversions and Markov workflows.

## Parent Wave
- Wave epic: `#33`
- Wave label: `wave:4`
- Milestone: `v0.4-structured-gp-and-uncertainty-aware-nn`

## Motivation
Temporal and state-space structure is the backbone of scalable spatiotemporal GP modeling: converting kernels to SDE/SSM form turns O(N³) GP inference into O(N) Kalman recursions, and it is what the Wave-5 integrations and the Bayesian-neural-field demo build on. This epic tracks the kernel-to-SSM conversion layer and the Markov GP workflow layer.

## Issues
- [x] #37 — gp(state-space): SDE kernels + composition rules (closed)
- [x] #38 — gp(markov): MarkovGPPrior + Kalman workflows (closed)

## Execution Notes
Both implementation children have landed. Remaining before closing this epic: verify the Definition of Done (clear representation/model separation, exports, one working temporal-GP example path) and either close the epic or attach any temporal-GP follow-up issues here.

## Parallelism
- Can run in parallel with: #35 and #36 once prerequisites exist
- Blocked by (inside this wave): none
- Must complete before: Wave 5 work that assumes temporal/state-space support

## Definition of Done
- SDE kernels and Markov GP workflows are implemented with clear separation between representation and model use.

## Relationships
- Parent wave: #33.
- Blocked by #30.