Regarding 301, the wiki states that "The duration of this spell is based upon the endroll but not in a simple manner. The minimum duration is 8 seconds (which is more likely to occur on a 101 endroll, but still may happen on a 200 endroll)."

However I have gotten several < 8 second stuns. For example in the following log the stun was ~4 seconds (enough to use one magic item):

You gesture at a n'ecare.
CS: +432 - TD: +323 + CvA: -6 + d100: +100 == +203
Warding failed!
A n'ecare shudders in mid-stride, then suddenly ceases all movement.
Cast Roundtime 3 Seconds.
(Forcing stance down to guarded)
>get orb from cl
You remove a heavy quartz orb from in your hooded black cloak.
>rub orb
You rub the quartz orb in your hand.
A sudden clarity of thought comes over you, and you realize that you now understand arcane mysteries that were previously incomprehensible.
Cast Roundtime 3 Seconds.
>put orb in cl
You put a heavy quartz orb in your hooded black cloak.
The n'ecare breaks free of the force that was holding it.

Without getting into questions of why it should even be possible to get a min 8 second stun on a 200+ endroll with a variable mana cost spell, is the wiki wrong or is the spell working as intended (there is no minimum stun)?

Thank you for the prompt response and confirming the min duration.

In the longer term, can the mechanics of this spell be reviewed? As you note the effective duration is only 4s since it takes 3s to cast the spell. Other disabling spells like Calm (201) have a flat benefit for warding margin, and AFAIK most Sorcerer spells have specific warding margin thresholds. Considering the mana cost increases based on target, it seems inconsistent with other stunning spells to have a 10+ mana spell be affected by a random component in ways other spells are not. Just my two cents.

Does it though? Breaking free of 301 is not a creature maneuver, and breaking free is due to "random chance" according to Estild. I have definitely been attacked right after breaking free (realizing that it is hard to show causation).

It depends. Prayer of Holding (301) is a persistent spell that triggers every 7 seconds. When it triggers, there is a random chance the creature can break out out of the binding and end the spell. The strength of the spell is based upon Cleric spell ranks and the warding margin of the initial cast. With a very high level cleric with a 200+ warding endroll, a creature will only have something like a 5-10% chance to break out when the effect first triggers (7 seconds after the cast). As time goes on, with each new attempt, the strength of the spell gets weaker, so the chance for the creature to break out increases (e.g. the creature might have a 5-10% chance to break on the first attempt, then a 10-15% chance to break on the second, then 15-20%, etc). This break attempt is a part of the spell itself and not a creature action.

However, some creatures innately break immobilization effects and for those creatures, a lot of time it does count as an action (so they usually can't immediately attack you). That's not always the case though and varies for each creature.

GameMaster Estild

Just confirming this spell is working as intended - the creature attempted to break out right after the spell was cast (not after 7 seconds).

Your spell is ready.
You gesture at a n'ecare.
CS: +424 - TD: +326 + CvA: +13 + d100: +97 == +208
Warding failed!
A n'ecare shudders in mid-stride, then suddenly ceases all movement.
Cast Roundtime 3 Seconds.
(Forcing stance down to guarded)
A n'ecare struggles against the forces holding him without much success.

Thanks for the prompt response Estild.