Stabalize Calibration network (calibnet)

## Background
At least as of 2025-11-25, [one SP has more than 50% of the power](https://calibration.filfox.info/en/ranks/power).

If that SP is temporarily down, the time to get messages on the chain takes longer. (This happened on 2025-10-28.)

We want to have enough power here that it’s non-trivial for non-implementation teams to fork the calibration network.

## Done Criteria
3 "just right" sized miner from 3 different organizations so there is enough of a moat while having some resiliency to one SP being down.

Specifically, this translates to:

- [x] Venus keep their miner `t04040` at ~250 TiB of QAP
- [x] StorSwift add their miner `[t0180698](https://calibration.filfox.info/en/address/t0180698)` with ~250 TiB of QAP
- [x] Curio reducing miner `t0143103` to 250 TiB of QAP
- [ ] Chainsafe bringing up a miner with 250 TiB of QAP. (Curio should be able to help seal 25 TiBs of sectors.)

As a bonus, since this is about calibnet stability, we'd like to have Probelab add monitoring for calibnet to https://probelab.io/filecoin

## Why Important
This is coming at a time where there is a lot of development and testing on calibration with Filecoin Onchain Cloud.

## Notes
1. This was discussed at FDS-7 ([session notes](https://docs.google.com/document/d/1kxA2b5V0Hm0lSF8ZQxmqauBoh6wJWXINscDpEm3sTH4/edit?tab=t.y3uj8nmj7dfk#bookmark=id.c5fo8a1oldav))
2. [Slack thread in #fil-net-calib-discuss](https://filecoinproject.slack.com/archives/C01D42NNLMS/p1764221330392239)