Som · 2 November 2026 · for Vin and Bobby · field names only · no export
Bobby's top 5: Hypertune PR Angles: Top 5 (31 Oct). This note does not confirm that a scan stores any of these fields. The facts page says a scan reads CPU, GPU, memory and storage. It does not list memory profile, power plan, bloatware, manufacturer, model or install date as saved fields. Austin or Kevin has to confirm the field list before W6-4.
Bobby's ask: rank problems by how often they appear in scans, not by FPS gain. Include the share of PCs that needed nothing. A Rust cut is separate. Publish anonymised totals only.
| Needed | Do we hold it? |
|---|---|
| A stored field per scan for each setting he wants to rank (memory profile, power plan, bloatware, and any other "wrong default") | Not confirmed. The product has Memory Optimizations and Windows Debloating, which is not the same as a saved yes/no on each scan. |
| Group size: how many PCs are in the scan set | Not in the facts page. Do not use "tens of thousands" until Kevin gives a count under the privacy floor in the method template. |
| The share of PCs that needed nothing | Only if the export has a row for every scanned PC, including ones with no change. Not confirmed. |
| A Rust cut of the same ranking | Needs a stored game or a separate Rust test. Game Hub can target a game. A scan field named Rust is not confirmed. |
| Gaps | No field list, no sample size, no privacy floor yet (see W7-1). Without those, this angle is not ready to pull. |
Bobby's ask has two parts. The scan part is existing data. The five borrowed PCs are a new test, not data we already hold.
| Needed | Do we hold it? |
|---|---|
| Manufacturer and model | Not confirmed as a stored scan field. Bobby says Windows reports them. That does not mean the Hypertune export keeps them. |
| Install date | Not confirmed as a stored scan field. |
| Memory profile on or off, and bloatware present | Not confirmed as stored fields. Same gap as angle 2. |
| Group size for "new" PCs versus all scans | Not held. "New" needs a rule (install date inside a window) written in the method before any cut. |
| Five borrowed new PCs, out-of-box, then free fixes, then Hypertune, tested in Rust | Not held. That is a study still to recruit. It is not in the scan archive. |
| Gaps | Do not rank brands from scans until the field list and the group sizes are confirmed. Report by setting, not by brand, until the groups clear the privacy floor. |
Angles 2 and 4 both wait on one answer: which of these names exist as columns on a scan export. Memory profile, power plan, bloatware, manufacturer, model, install date, and a flag for "no change". Until that list comes back, neither angle should be pitched as ready.