v0.16.0 - Experimental Windows release

TrueTrack v0.16.0 — understandable results and fixed baseline

Run TrueTrack.exe after extracting the complete ZIP. Preserve your existing profiles folder when updating.

• Results explain what improved, what regressed and which safeguards blocked the recommendation, followed by gauges.
• A fixed snapshot of the first baseline is retained through later rounds. A/B and first-baseline progress are separate; incompatible target protocols are clearly marked.
• New rounds use balanced near/far targets and fixed directions in randomized order. Equal target counts have equal total planned distance; A/B always has identical paired geometry. Adaptive rounds still compare per-target statistics, not raw totals with unequal counts.
• New rounds can support a profile through either faster precision hits or clearly lower corrective travel, with hit/time/flick safeguards. The two precision routes split their statistical threshold; a green gauge is not proof.
• Optional circle verification records 30 clockwise and 30 counterclockwise attempts with the already active Raw Accel proposal, then overlays its measured median against the first baseline. No refitting or automatic activation.
• Center-the-mouse instructions before tests. English is the first-launch default; existing language preferences remain.
• Current-only ALGORITHM.md documents the exact protocol, thresholds, baseline storage and measured circle verification.

Old measurements and frozen confirmations keep their original assessment and target plan. Start a new calibration or next phase to use the new protocol. No private measurements are included. Experimental unsigned Windows x64 build; optional global driver/game compatibility remains game-dependent.

Validation: 400 automated checks passed; no driver installed or settings activated by these checks.