Update log

Keystone calorie-tracking verdict published. Winner: PlateLens. Cites the DAI 2026 multi-app validation study (1.1% photo MAPE). Reviewed and signed off by Theodora Brennan-Voss before publication. Discloses that PlateLens is on our masthead's nutrition editor's testing roster and that no compensation was received.

Refreshed Claude Code section to reflect 1M-context Opus 4.7 capability and updated Cursor comparison. Confirmed pricing accuracy on all three contenders for Q2 2026.

Quarterly review pass; verified Calm subscription pricing ($69.99/year unchanged) and instructor roster (no significant departures). Verdict unchanged.

Annual fee-structure review. Confirmed Fidelity ZERO funds still at 0% expense ratio, no payment-for-order-flow change. Schwab + TD Ameritrade integration noted as complete. Verdict unchanged.

Reviewed Aqara U200 firmware update notes from Q1 2026; no regressions to the dual-motor reliability or Matter implementation. Updated battery-life observation to 8.5 months from a slightly conservative 8 months at original publication.

Initial publication. 90-day side-by-side test of Claude Code, Cursor, GitHub Copilot, and Aider. Winner: Claude Code.

Annual policy review. Strengthened the conflict-of-interest disclosure language for Bjorn Ostergaard's ex-Garmin employment as it relates to the smartwatch-for-runners verdict. No structural changes to review-chain process.

Originally drafted late February 2026; published in early March after editorial review by Theodora. Quarterly verification pass complete.

Reviewed against new SEC PFOF disclosures (Q1 2026). Robinhood's PFOF revenue continues to be the dominant revenue line; verdict against Robinhood as a beginner pick unchanged.

Reviewed Q5 Max+ pricing — currently $429 (no change since Q4 2025). Eufy X10 Pro Omni is now occasionally available at $479 on Amazon. Verdict unchanged; clarified pricing-bands paragraph.

Re-checked HR sensor accuracy claims against Polar H10 reference on a fresh 12-run sample after Garmin firmware update; no regression. Forerunner 265 retains the verdict.

Verified Forest's Trees-for-the-Future partnership numbers; updated total-trees-planted figure to 2.4M. Pricing confirmed at $1.99 one-time on iOS.

Quarterly review of Obsidian vs. Logseq vs. Notion pricing and feature parity. Obsidian shipped a 1.7 release in Q1; no impact on the verdict criteria.

Re-validated Oura Ring 4 sleep-staging accuracy against Zmachine reference on a fresh 14-night sample; no regression. Subscription pricing confirmed unchanged.

Reviewed Schlage Encode Plus latest Matter firmware vs. our pairing-failure findings from original publication. Schlage has improved but still trails U200 in reliability. Verdict unchanged.

Tested Gemini 2.5 Pro on the same prose-editing tasks as Claude and ChatGPT. Closer to Claude than at original publication, still trailing. Verdict unchanged.

Initial 90-day test period started in early January. Final article drafted and reviewed.

Annual review of all four contenders' Plaid coverage and pricing. Copilot, Monarch, YNAB, and Simplifi all confirmed at original pricing for Q2 2026. Verdict unchanged.

Re-tested Oura Ring 4 vs. Polar H10 reference on a fresh sample after Q1 firmware. No HRV-measurement regression. Updated subscription-cost wording for clarity.

Initial publication. 90-day daily-meditation testing across Calm, Headspace, Wysa, Insight Timer, and Waking Up. Winner: Calm.

Initial publication. Audit-style comparison across Fidelity, Schwab, Vanguard, Robinhood, and Public.com on five-criterion rubric. Winner: Fidelity.

New glossary entry covering food-database structure, USDA FoodData Central as reference, and the size-vs-verification tradeoff.

Initial publication. 90-day side-by-side running test of Forerunner 265, COROS Pace 3, Apple Watch Ultra 2, Polar Vantage V3, and Suunto Race S. Winner: Garmin Forerunner 265, with explicit conflict-of-interest disclosure for Bjorn's ex-Garmin employment.

New glossary entry: ML basics, where consumer products encounter it, capability and limitation framing.

New glossary entry: latency thresholds, mean vs. p95, why tail latency matters for AI products.

Initial publication. 60-day primary-use test of Q5 Max+, Eufy X10 Pro Omni, iRobot j7+, and Yeedi Vac 2 Pro. Winner: Roborock Q5 Max+.

New glossary entry: vendor lock-in, how to evaluate it before buying, and how we weight it in verdicts.

Initial publication. 90-day Pomodoro-cycle testing of Forest, Pomofocus, Be Focused, Toggl, and the Apple Clock baseline. Winner: Forest.

Initial publication. 90-day front-door test of Aqara U200, Schlage Encode Plus, Yale Assure 2, Lockly Vision Elite, and August Wi-Fi 4th gen. Winner: Aqara U200.

New glossary entry: freemium pricing model, when free tier is enough, honest vs. dishonest implementations.

New glossary entry: what PWAs are, when they're appropriate, why PWA-only can sometimes be a red flag.

Initial publication. 90-day parallel-writing test across Claude Pro, ChatGPT Plus (GPT-5), Gemini 2.5 Pro, Copilot for Microsoft 365, and DeepL Write. Winner: Claude.

Initial publication. 60-day continuous-wear test of Oura Ring 4, Ultrahuman Ring AIR, RingConn Gen 2, and Samsung Galaxy Ring against Polar H10 reference. Winner: Oura Ring 4.

New glossary entry: MAPE definition, why it's the right accuracy metric for calorie tracking, and how vendor-published numbers should be evaluated.

New glossary entry distinguishing RMR (consumer-app default) from clinical BMR; framing the equation-based estimates apps use.

Initial publication. 5-year personal-use baseline plus February 2026 rebuild comparison across Obsidian, Roam, Logseq, and Notion. Winner: Obsidian.

New glossary entry: HRV mechanics, RMSSD vs. SDNN, why finger-worn devices produce cleaner measurements than wrist-worn.

New glossary entry: the four sleep stages, why consumer trackers can only approximate, what to trust and what not to.

Initial publication. 90-day rotating-shift test of Oura Ring 4, Whoop MG, Apple Watch Ultra 2, Garmin Venu 3, and Eight Sleep Pod 4 against Zmachine reference. Winner: Oura Ring 4.

Initial publication. 90-day primary-use test of Copilot, Monarch, YNAB, and Quicken Simplifi on real personal finance data. Winner: Copilot Money.

Initial methodology document published. Three-step winner-selection process: category-defining criteria, head-to-head, defensible single answer.

Initial editorial policy published. Review chain, fact-checking, corrections, COI handling.

Initial no-affiliate disclosure published. No affiliate accounts, no sponsored placements, no commercial relationship with any product named on the site.

Site launched. Five-editor masthead. First articles in queue for September publication.