This is a useful corrective to threat-inflation, and it should be read against today's news, not in isolation. Blair is right that an untested force is an unknown quantity, and that the PLA has not fought a real war since 1979. But the corrective has its own failure mode: comfort. A Liaoning group completing a 40-day deployment and four warships facing down a Philippine frigate are not hardware-count abstractions; they are reps, and reps are how an untested force becomes a tested one. The honest net assessment is not "China is a paper tiger" or "China has won"; it is that the gap is closing through exactly the sustained operations Blair correctly notes are unproven. Clausewitz's point about the difference between war on paper and real war cuts both ways: it humbles Chinese confidence, and it should humble ours about predicting how their inexperience plays out. Worth reading alongside the Scarborough item to keep the abstraction honest.

Kofman is the analyst to read on Russian military power, and the thesis lands because it inverts the comfortable story. The frozen front and the fuel-sales halt in Crimea, real as they are, measure today's Russia, not tomorrow's. Armies that survive a hard war often come out more lethal, the Wehrmacht after 1940, the US Army after Vietnam's hollowing and rebuilding, because combat is the most expensive education there is and Russia is paying full tuition. The strategic error the West keeps making is linear extrapolation: assuming the Russia of 2026 attrition is the Russia of 2030. This connects directly to the burden-sharing fight: if Hegseth's European drawdown assumes a permanently degraded Russia, and Kofman is right that reconstitution is the real threat, then the drawdown is timed to the wrong curve. Read this and the Germany-strategy piece together. The question both raise is whether Europe rearms faster than Russia rebuilds.

The contrarian framing is sharper than it first looks. The Ukraine-driven consensus, "drones change everything, buy more," skips the harder question the piece foregrounds: a weapon that is cheap for you is cheap for the enemy, so a drone-saturated battlefield advantages whoever solves the counter-drone problem, not whoever fields the most drones. That is the right lens. The Crimea strikes prove drones can impose cost; they do not prove drones win wars, because Russia is absorbing the cost and the front has not moved. The doctrinal trap is treating a leveling technology as an edge. The MWI argument maps onto the materials story too: if counter-UAS depends on interceptors and electronics that run through the same strained, China-throttled supply chain as everything else, then "solve the drone problem" runs into the same industrial-base wall as refilling the missile magazines. The future of war is not the drone; it is who can produce the counter at scale. Worth the full read for anyone building a force-design argument.

This is the document that ties the whole brief together, and it is worth reading precisely because it refuses the comfortable abstractions. Every other thread today, the munitions math behind the Iran deal, the rare-earth squeeze, the three-theater budget straining behind the new Pacific command, is a symptom of the single condition the estimate names: plans have outrun logistics. The value of an institutional baseline like this is that it converts the anecdote into a requirement, "we are short" becomes "we are short against these 105 command-sponsored requirements," which is the language that actually moves appropriations. The Clausewitzian point sits underneath it: friction is not the exception in war, it is the medium, and an industrial base that cannot replace losses turns friction into culmination. The honest read is that the United States has spent a year discovering its strategy assumed a logistics system it does not have. Whether the FY2027 NDAA closes that gap or just names it is the question the estimate frames and the appropriators will answer. Cross-referenced from institutional research feeds; cite the War College Press monograph directly.