Or else whichever claimant wins will undo the CA penalty immediately on accession as long as he has enough votes and isn't particularly hated by his vassals. A chain of quick successions can also fast-track the realm from e.g. Autonomous Vassals to Medium or from Medium to Absolute, especially given the opinion penalty for increasing crown authority is not heritable, meaning that the successor is not affected by it, and starts with a clean slate.
You may disagree with me but I believe it would make more sense to restrict crown authority changes heavily than to do so with some of the changes to succession law. From the point of view of a vassal not belonging to the royal dynasty or not having either a strong position in the line of succession or the means to support his claim forcibly, taking away his right to conduct wars and messing with his own inheritance (inheritance outside the kingdom) should matter much more than whether the succession is Agnatic Primo or Agnatic-Cognatic Primo, or, in many cases, primo vs gavelkind (where there is just 1 top title, especially given that vassals tend to care mostly about the de iure title they belong to, rather than the entire de facto realm).
For example, suppose I'm the Duke of Toulouse under France. I don't care whether the King will hand out counties and duchies to his sons when they turn 16 or when he dies, or whether the crown will go to his brother or daughter. But I do care when I'm forbidden from meddling with my neighbours, let alone from participating in holy wars in Spain, or crusades, or effectively precluded from sharing inheritance with my cousins in Provence on the Imperial side (or Spanish lordships).
In the long term, I believe it could make sense to split the various competences of the kings and emperors and the restrictions on nobles into individual decisions. If you think about it, it is kinda strange that if you want your vassals to stop in-fighting, you can only do it in bundle with increasing the size of the levies they're supposed to provide. We could even keep a 'Crown Authority' index calculated as a (weighted) average of such individual decisions--thus effectively a shorthand. But this is just speculation. In the short term, I do think restricting the AI from rushing to higher crown authority immediately after winning a civil war to depose the liege or install a claimant would be a good solution. (As well as cutting off too significant impact of insignificant reigns.)
You may disagree with me but I believe it would make more sense to restrict crown authority changes heavily than to do so with some of the changes to succession law. From the point of view of a vassal not belonging to the royal dynasty or not having either a strong position in the line of succession or the means to support his claim forcibly, taking away his right to conduct wars and messing with his own inheritance (inheritance outside the kingdom) should matter much more than whether the succession is Agnatic Primo or Agnatic-Cognatic Primo, or, in many cases, primo vs gavelkind (where there is just 1 top title, especially given that vassals tend to care mostly about the de iure title they belong to, rather than the entire de facto realm).
For example, suppose I'm the Duke of Toulouse under France. I don't care whether the King will hand out counties and duchies to his sons when they turn 16 or when he dies, or whether the crown will go to his brother or daughter. But I do care when I'm forbidden from meddling with my neighbours, let alone from participating in holy wars in Spain, or crusades, or effectively precluded from sharing inheritance with my cousins in Provence on the Imperial side (or Spanish lordships).
In the long term, I believe it could make sense to split the various competences of the kings and emperors and the restrictions on nobles into individual decisions. If you think about it, it is kinda strange that if you want your vassals to stop in-fighting, you can only do it in bundle with increasing the size of the levies they're supposed to provide. We could even keep a 'Crown Authority' index calculated as a (weighted) average of such individual decisions--thus effectively a shorthand. But this is just speculation. In the short term, I do think restricting the AI from rushing to higher crown authority immediately after winning a civil war to depose the liege or install a claimant would be a good solution. (As well as cutting off too significant impact of insignificant reigns.)