Should Wiktionary avoid indefinite blocks of productive users?

In the English Wiktionary, indefinite blocks of productive users sometimes happen. Should this be avoided? An obvious alternative is to cap blocks at e.g. one year.

Predicates:
 * pure grafiti-like vandal, pure spammer, productive user, productive user with long-term contribution record, user who causes drama and exhausts people in non-productive discussions, user who insults users, user who insults users with vulgarities, user with long track-record of adding unattested entries, user with long track-record of too many inserted errors
 * one-day block, three-day block, one-week block, one-month block, two-month block, three-month block, six-month block, one-year block, indefinite block, block from a specific namespace.

Arguments for

 * An indefinite block is un-Christian, failing to implement Christian forgiveness.
 * Wiktionary is not specifically Christian project. It is not clear why the Christian values of love, kindness and forgiveness should apply.
 * At least the English Wiktionary does tend to be more Christian in the cultural background, given the cultures in English-speaking countries; even the atheists often have something like Christian cultural background. At least there, it does make sense to think of Christian values.
 * The above comment applies to various Wiktionaries of European languages as well.
 * A one-year block is enough to signal something is very wrong and to prevent disruption over a considerable period of time. It should be straightforward for qualified administrators to block bad behaviors as they happen, whenever they happen. Even if the block candidate editor has no chance of changing their behavior--rather improbable in general--this would produce administration overhead with the frequency of once per year only.
 * An indefinite block causes a loss of contribution by the blocked person while not being the minimum necessary measure to prevent project disruption.
 * Some contributors can have a mild form of mental illness, which may explain some of their bizarre behaviors. An excellent and professional team of administrators manages such a person as an asset, with adequate kindness and forgiveness, rather than as a mere nuisance, to maximize the value for the project in the form of contributions. Some people with mild mental illness may have disability pension, which would make it possible for them to devote a lot of time to the project. For instance, the English Wiktionary had a somewhat more disputatious person from the U.K. who reported to have a depression or something of the sort, and while he was arguably too passionate in certain disputes, he was a great and valuable contributor.

Arguments against

 * Indefinite block is fine as long as the user is able to request an unblock, declare what mistakes they recognize and indicate what changes in behavior they plan.
 * This is unnecessary, especially given that some users keep on promising things that they do not abide by anyway. It is all too often a game, not a genuine thing. One case in point is an English Wiktionary editor who edited in many languages, inserted too many mistakes, kept on promising things again and again, but kept on breaking the promises.
 * It does not need to be a mere game. In any case, there is something psychologically significant about person publicly recognizing their mistakes and being able to formulate changes to future behavior.
 * It is actually likely to be a game, where the blocking administrator derives illicit pleasure from humiliating another human, and exercising often rather arbitrary power over them.
 * Sometimes, a disruptive person's behavior is not a clear-cut case for block. Then, when a clear-cut misbehavior happens, it is essential to block the person indefinitely since the less clear-cut bad behavior may happen in future and be hard to block.
 * That approach to blocking is Machiavellian, dishonest and should not be tolerated or supported.
 * There are cases that perhaps do justify an indefinite block, e.g. making threats of violence to users, doxxing of physical address of residence, or similarly egregious behavior.
 * Indefinite blocking of obvious vandals is fine, and causes no loss of contribution.
 * The motion is about productive users, not pure vandals.
 * But it is unnecessary, and there can be a loss of contribution in principle since an editor who starts as a vandal (perhaps a teenager in puberty showing typical puberty behavior) can turn into a productive editor later.
 * In many wiki projects, anonymous vandals only get a very short block or none at all and get a warning instead. That has proved to work reasonably well. For a repeated vandal, issuing a one-year block should be enough. Not conclusive, yet suggestive.
 * Indefinite blocking of obvious spammer account is fine, and causes no loss of contribution. Spammers are very unlikely to turn into productive users; they are not mischievous teenagers but rather instrumentally rational agents, often with monetary incentive.
 * The motion is about productive users, not spammers.