This low-dimensional thermally driven model, run as numerical integrations rather than ensemble statistics, demonstrates that slow periodic forcing of the alpha-effect produces a multimodal persistence-time distribution whose peaks align with integer multiples of the modulation period. Unlike full 3D geodynamo simulations that treat reversal statistics as purely internal chaotic noise, the setup isolates how external pacing can impose regularity on an otherwise broad-tailed waiting-time distribution. The work is a 2026 arXiv preprint and therefore lacks peer review or independent code verification; its dimensionality reduction necessarily omits small-scale turbulence and heterogeneous boundary conditions that real-Earth paleomagnetic records suggest matter. Related analyses in a 2008 GRL study by Constable & Korte on paleomagnetic reversal clustering and a 2021 PRE paper by Pétrélis et al. on stochastic resonance in low-order dynamo maps already hinted at similar resonance windows, yet neither imposed explicit thermal modulation. The present results therefore close a mechanistic gap by showing that the observed superchron-to-short-reversal spectrum can emerge from a single control-parameter rhythm without invoking separate dynamo regimes. Limitations include the absence of reported ensemble size, sensitivity tests on modulation amplitude, and direct comparison against the most recent GPTS timescale.