What the Spec Sheet Leaves Out
A gravel bike spec sheet tells you the frame material, the groupset, the wheel size, and a number for max tire clearance, usually somewhere around 45mm. It doesn't tell you what happened when someone in the design shop mounted a 50mm tire and the chainstay wouldn't clear the crank arm at full compression, or the seatstay flexed enough under load to rub the tire on every third pedal stroke. The clearance number on the sheet isn't a target they picked and hit. It's a boundary they found by going past it first.
Read enough spec sheets and the pattern holds everywhere. "Not compatible with a front derailleur" isn't a design philosophy stated for your convenience, it's an admission that someone tried to route a front derailleur cable through that frame and the geometry fought back. "Requires additional sealant tape for tubeless" means the stock rim bed leaked air on the bench and nobody wanted to redesign the rim to fix it. "Max rider weight 275 lbs" is a fatigue test that ended before the frame did, at some specific cycle count someone doesn't want to print. Every one of these lines reads like a rule handed down from principle. Every one of them is actually a scar with the story sanded off.
The feature list works the opposite way. It's written forward, toward a sale, and it has every incentive to round up. Twelve-speed drivetrain, tapered head tube, internal cable routing: all true, all chosen because they sound good on a page someone is scanning in a shop. Nobody pads a feature list with things that turned out badly in testing. So the feature list is honest in the narrow sense that the claims are usually verifiable, but it's curated the way a resume is curated. It tells you what they're proud of finishing.
The exclusion list has no such incentive. Nobody adds a fake limitation to a spec sheet to move units. If anything the sales pressure runs the other direction, toward quietly dropping a constraint that makes the product look worse. So when a limit survives onto the printed sheet anyway, that's a signal that the failure behind it was expensive enough, or embarrassing enough, or dangerous enough, that engineering won the argument with sales. The 45mm ceiling stayed on the page because letting a customer run a 50mm tire and snap a chainstay was worse than losing the sale to a competitor's bigger number.
This isn't unique to bikes. An API's rate limit is the record of some client that hammered an endpoint hard enough to page someone at 3 a.m. A footnote that says a flag "is not supported in combination with" another flag is two engineers finding out the hard way that the combination corrupts state, then writing the incompatibility down instead of fixing the underlying interaction, because fixing it was harder than fencing it off. An electrical panel's "not rated for continuous load above 80%" is a breaker that tripped, or worse, one that didn't. In every case the excluded thing used to be possible. Somebody made it stop being possible on purpose, and the spec sheet is where they filed the decision.
None of this means the feature list is lying. It just means it's describing the part of the product that behaved. The exclusion list describes the part that didn't, and got contained instead of solved. If you want to know what a piece of engineering is actually made of, don't read what it does. Read what it flatly refuses to do, and ask what it cost someone to learn that refusal was necessary.