Plan an Incremental UI Migration

Background: Native [iOS / Android] app currently on [UIKit / Android
Views / a mix]. We want to move to [SwiftUI / Jetpack Compose] without
stopping feature work. App size: [number of screens]. Minimum OS:
[version]. Team size: [n]. Release cadence: [e.g. every two weeks].

Use case: Give me a migration plan we can execute a slice at a time,
shipping the whole way, rather than a rewrite we'd have to defend for
six months.

Implementation details:
- Interop is the crux: how [UIHostingController / ComposeView] and the
  reverse direction work, and where the boundaries should sit
- Which screens to convert first—optimise for low risk and high learning,
  not for the screens we dislike most
- What to do about navigation, which usually can't be migrated piecemeal
- Shared state and theming across the old/new boundary
- The point at which a half-migrated app costs more than either end state,
  and how to avoid stalling there
- Which screens might be right to never migrate

Deliverable: A phased migration plan with the interop pattern for each
direction, an ordered screen list with the reasoning, the navigation
strategy, the specific risks at each phase, and the exit criteria that
tell us a phase is genuinely done.