Multi-year comparisons of self-hosting versus cloud subscriptions rarely account for manifest-sourced price histories, storage-tier jumps, hardware replacement, and explicit rules for licensed media. We backtested nine synthetic Brazilian household configurations across 2015–2023 entry years using 103 subscription observations, 718 self-hosted cost rows, and a 540-cell deterministic grid (four media scenarios, IPCA-deflated real prices). Under conservative media (streaming not counted as avoidable), focal entry 2021 with a 5-year horizon yields positive net savings only for storage-heavy personas: P2 Google/Apple (+R348 to +R541), P3 Google (+R1, 170), and P4 (+R4, 975) ; P1 and streaming-only P5 never reach payback. Identical NVR hardware produces opposite P3 outcomes when camera cloud is itemized (Google) versus bundled in iCloud+ (Apple) ; a ±20% hardware capex perturbation flips P2 Google focal net from +R918 to−R222, joint hardware–energy stress (+20%/+20%) turns the same cell negative, and literature-interval media-fraction bounds span ≈R1, 800 on P5ₕybrid focal net. Practitioners should model multi-service baskets and tier jumps, not per-gigabyte list prices alone; labor, reliability, and privacy sit outside the cash-flow model.
William Oliveira (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: