[KV Feedback]: Replication delay / visibility

### 🔍

- [ ] Did you search for existing issues?

### Type of feedback

General feedback

### Description

I keep running into situations where someone sees different results from a Deno Deploy server, specifically the KV values.

It's frustrating - I can't tell what's happening, why it's the case. All we know is if we do requests from different regions (using a VPN) we get different results.

Presumably it's because values are propagating through the various KV servers ... when I look at the logs I can see it's certain machines ... but I can't tell where, or why, or for how long ...

I would really appreciate an ability to control this, or at least gain some visibility somehow. We've long since turned off all the write/read regions we could, for this very issue, but it hasn't helped.

Not really sure what I'm suggesting. A hard flush? How does propagation work? Foundation DB? Can we have access to some of those controls?

Anyway. Just wanted to write a note. Thanks for reading.

### Steps to reproduce (if applicable)

_No response_

### Expected behavior (if applicable)

_No response_

### Possible solution (if applicable)

_No response_

### Additional context

_No response_

No files, tests, or entry points are named. Start by reviewing the reported cross-region KV discrepancies and replication-delay questions; the issue does not define a specific change or a clear completion condition.

Written by the indexing model from the issue text.