Background/Objectives: Diagnostic criteria for non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) have changed twice since 2016, most visibly in the tolerance for true papillae and in the role of molecular findings. We evaluated historical NIFTPs under the original and subsequent frameworks and related framework-specific classification to long-term outcomes. Methods: In this single-center, retrospective cohort (surgery 2007–2016), archival slides of 79 historical NIFTPs with linked longitudinal follow-up were independently re-reviewed in 2026 by two pathologists without knowledge of long-term outcomes or molecular results; their framework-specific classifications were concordant in all cases. Reasons for failure to retain the diagnosis across all frameworks were classified as criterion-dependent, criterion-independent, or technical. Outcomes were compared with 56 infiltrative and 39 invasive encapsulated follicular-variant papillary thyroid carcinomas (FVPTCs) from the same institution. Results: Thirty-six tumors (45.6%; 95% CI, 34.3–57.2%) qualified as NIFTP under all frameworks. Retention was 47/79 (59.5%; 95% CI, 47.9–70.4%) under the original 2016 criteria, 37/79 (46.8%; 95% CI, 35.5–58.4%) under the 2018 consensus primary criteria, and 46/79 (58.2%; 95% CI, 46.6–69.2%) under WHO 2022. Ten tumors with well-formed true papillae occupying <1% qualified under 2016 and WHO 2022 but not under the 2018 criteria. One additional BRAF V600E-positive tumor had no well-formed papillae, an entirely follicular growth pattern, and a nuclear score of 3; it fulfilled the 2016 and 2018 primary morphologic criteria but was excluded under WHO 2022 because BRAF V600E was detected. Criterion-independent discordance between the historical classification and contemporary re-review accounted for 28/79 tumors (35.4%; 95% CI, 25.0–47.0%). During a median follow-up of 132 months, no incident structural events were observed among historical NIFTPs (0/64; upper 95% CI, 5.6%), and none of 36 with examined lymph nodes had nodal metastasis, whereas nodal metastasis at surgery (4/27) and the only incident structural recurrence occurred among invasive encapsulated FVPTCs. Conclusions: Framework-specific retention ranged from 46.8% to 59.5%, driven mainly by the papilla criterion and, in one molecularly discordant case, BRAF V600E status. Most diagnostic loss reflected criterion-independent discordance identified on contemporary dual-pathologist slide re-review. No incident structural events were observed among outcome-evaluable historical NIFTPs during a median follow-up of 11 years.