Imagine you have a shelf of books sorted by size, tallest on the left. Someone comes along and swaps books around — now the tall ones are scattered across the shelf. That disorder IS stored energy: you could extract work by letting gravity pull each book back to its sorted position. The amount of work you could extract depends on how far from sorted the shelf is, not on whether the books are vibrating. That is ergotropy — extractable work stored in a non-passive arrangement — and this paper shows that a two-qubit quantum battery stores its useful energy almost entirely through this population reshuffling mechanism, not through quantum coherence. The committed claim: in a capacitively coupled double-quantum-dot (DQD) battery charged by resonant microwave tunnel-coupling modulation, the dominant extractable work (ergotropy) is stored in non-passive population distributions of the energy eigenstates, not in quantum coherence. This is the Nth iteration of the population-vs-coherence ergotropy debate in quantum batteries, but the first to do it with experimentally motivated 1/f charge noise on a semiconductor DQD platform. The architecture is clean and specific. Two charge qubits coupled via an Ising-type capacitive interaction, driven by resonant modulation of the tunnel coupling. Detuning fluctuations enter as classical stochastic 1/f noise drawn from a band-limited spectrum. Each noise realization evolves unitarily — decoherence only appears after ensemble averaging over many realizations. The authors decompose total ergotropy into a population contribution (from non-passive ordering of energy-basis populations) and a coherent contribution (from off-diagonal elements), then track both under noise. The key finding: resonant driving selects a dominant E0-to-E3 population transfer channel, and this channel carries most of the extractable work. Coherence is a transient dynamical companion, not the load-bearing structure. On the ladder, this paper sits within a growing body of quantum battery theory — Alicki and Fannes (2013) introduced ergotropy-based QB analysis, and recent work by Andolina et al., Ferraro et al., and others has explored population-vs-coherence decompositions. The specific contribution here is pairing the population-ordering interpretation with a realistic noise model (1/f charge noise) on a semiconductor platform rather than idealized open-system Lindblad dynamics. No named numerical baselines are beaten — this is a theoretical framework paper, not a performance benchmark. The paper does not compare its extractable-work figures against competing QB architectures or claim superiority in energy storage density. Integrity is the weak link, honestly. The entire validation is numerical simulation by the same team. The 1/f noise model is experimentally motivated — band-limited power spectral density matching measured charge noise in semiconductor quantum dots — but there is no experimental data, no independent simulation, and no community benchmark for quantum battery performance. The ensemble averaging approach (unitary per realization, decoherence after averaging) is physically well-motivated but makes the results dependent on the noise sampling procedure. The passive ordering violation diagnostic and Jensen-Shannon coherence measure are well-chosen analytical tools, but the paper is grading its own homework. The milestone question exposes the real gap. Quantum batteries remain firmly theoretical — no device has demonstrated net extractable work exceeding charging losses at the single-device level. The paper does not provide a concrete next number (e.g., a target coupling strength, noise floor, or fidelity threshold) that would bridge from simulation to a working DQD battery. The field needs someone to build one, measure the ergotropy decomposition experimentally, and see if population dominance survives real hardware noise. The obvious next experiment is running this on actual silicon or GaAs double-quantum-dot hardware with measured 1/f noise spectra and extracting the ergotropy decomposition from real charge-state tomography. The authors didn't do it — most likely because they are theorists without fab access (option a: resource constraint). A collaboration with an experimental semiconductor-qubit group would be the natural next step, and the framework is set up to make that comparison straightforward.