Shay

There is a real downside beyond electricity, and it is the one nobody mentions: colder air holds less moisture, so a colder freezer dries food faster.

Freezer burn is not the food freezing too hard, it is water leaving the food and refreezing on the packaging as frost. The drier the air in the cabinet, the stronger that pull. At −25 in a frost-free unit — which deliberately cycles warm to defrost, then re-dries the air — poorly wrapped food will develop freezer burn noticeably faster than the same food at −18.

So the answer to question 1 is: the cost is electricity plus a mild penalty on anything not sealed properly. In a chest freezer without auto-defrost this is much less of an issue than in an upright frost-free one, so your setup is the forgiving case.

Question 2 is the sharpest thing you asked and the answer is yes, they mean different things.

Gradually lengthening is wear. Something is degrading in a consistent direction — contacts pitting, a spring weakening, a bushing wearing oval. It will keep getting longer, on a schedule you can now roughly predict, and it ends in it not starting at all.

Intermittent — fine for a week, terrible one morning, fine again — is usually a connection that moves. Corrosion that conducts differently when it is damp, an earth strap that is tight but not clean, a connector that has backed out slightly. Those correlate with weather and with going over a bump, and they can go on for years.

Yours is the first kind, so I would not leave it. Not urgent this week; do not take it on a long trip in six months either.

On question 2, the kind of driving matters more than the hour does.

At idle or crawling in traffic, the alternator is spinning slowly and the car is also running the biggest loads it has — headlights, heater blower, heated rear screen, wipers. In winter, in traffic, at night, it is entirely possible for the car to be consuming everything the alternator makes and taking the remainder from the battery. An hour in that situation can leave you slightly worse off than you started.

An hour at steady speed with the lights off and the heater low is a completely different hour. Same clock, several times the charge.

So the advice should be "drive for an hour on an open road", and the version most people follow is "sit in traffic for an hour", which is why it sometimes appears not to work at all.

It is a one-way friction device, and the one-way part is the whole trick.

The derailleur cage is sprung to pull the chain tight. Over bumps, the chain slackens for an instant and then snaps back, which is the slapping you hear and, occasionally, how chains come off. The clutch adds friction that resists the cage being pulled forward — the direction it moves when the chain goes slack and then whips — while adding much less resistance to the cage moving backward, which is what happens when you shift onto a bigger sprocket.

So it is not a brake in both directions. If it were, shifting would be genuinely awful. It is closer to a damper that is stiff one way and soft the other, usually a wrapped band or a roller clutch on the cage pivot.

That asymmetry is why you can leave it on permanently, which is what it is designed for.

Worth adding the case where weight does matter, so the physics does not get dismissed entirely: stop-start city driving.

Every acceleration from rest is spent moving the mass, and none of it comes back. A car doing constant short urban trips will show a weight effect several times larger than a motorway car, because it is paying the acceleration cost over and over.

It is still small. But if someone wanted to be rigorous about it, half-filling a tank makes marginally more sense for a car that only does school runs than for one that does 200 mile motorway trips — which is the exact opposite of how most people think about it, since the motorway driver is the one who fills up religiously.