Does a magical focus only work for spells dedicated exclusively to the field of the focus, or can it's bonus apply based on the target even if not all targets of the spell would qualify?
For instance, if someone has a Major Focus on Spirits, and they cast Ring of Warding against Spirits, obviously the focus applies to the casting total to affect any possible spirit. But what if the magus in question casts Circular Ward Against Demons, which specifically wards against anything with Infernal might? Demons and Infernal Ghosts would fall under the purview of the focust, but an infernally currupted stoat or infernally animated corpse would not. Does the mage calculate casting total / penetration one way for the first example, another way for the second?
I would calculate casting total and Penetration differently for both, if the Focus could apply. Obviously for someone casting a Ward Against Beasts of Legend, a Magical Focus of Spirits wouldn't matter.
You can only have one Casting Total for a spell, determined at time it is cast.
Since the Casting Total determines if the casting succeeds at all, among other things, it can't have more than one value.
So for your example, with a spell that does not inherently have anything to do with your focus at the time of casting, the Casting Total would be calculated without the focus.
Penetration is normally based on the Casting Total, and that part of the calculation should be constant, but if you want to have "focus casting total" re-calculated for Penetration purposes only if something covered by your focus happens to be affected by the spell, I don't see a major problem with that.
I do it like Itzhak. I'm playing a sahir in a game, so if I cast a personal ward vs realm, I roll once, and write a different number for jinns vs non-jinn.
As someone currently playing a magus who has a Corpus focus in Men, I calculate for both if I'm round the table. If I'm using Foundry VTT, I ask the SG if my focus applies before ticking the box when rolling.
I prefer that Magical Foci be described in terms of themes of magical practice and/or effect and not in terms of targets.
I'm not hardline about it as it might be overly limiting for certain Focuses, but as a broad generality I prefer that spells you want to cast with your Focus be Focus-applicable for lab invention. This avoids such ambiguities and is often more flavorful due to the resultant spell's specificity. Like a Focus: Men maga interested in controlling and menacing men, warding her space so that no men can intrude upon her nor escape her captivity, but whose male-specific spells leave her vulnerable to the men's wives mounting a rescue mission... or something.
That said, that's a personal preference. As a rule, I think a target-based Focus only applies on casting if all applicable targets fit the Focus. With the aforementioned Focus: Men, you'd surely get it while casting a Group Corpus spell on the battlefield against an all-male battle group, but would have a harder time finding a Group you could get the bonus casting against in a bustling city of intermingling genders. This is mildly noisome because there's probably situations where you could gain information from your Focus unexpectedly failing to apply, which is another reason I prefer the "such spells should be invented to affect that specific type of target" approach.
Note that the "Focus provides information" problem occurs regardless of whether one allows broad spells to benefit from a narrow Focus when cast on appropriate targets (e.g. a Corpus spell that makes any human taller, but gains a bonus from your Focus on Men only if cast on a Man), or one allows spells to be restricted at the time of invention from working on anything outside of a Focus' scope (e.g. a Corpus spell that only makes Men taller).
That said, I do agree that the second choice is a far cleaner solution overall, and one that better balances the Magical Focus Virtues, particularly the Minor one.
You're right that it provides information either way, but imo it tends to do so more interestingly in the case where a spell unexpectedly fails in its entirety, especially if such doesn't make itself apparent until after a negative consequence has occurred.
Does a wizard actually know what their Casting Total is? It seems like they should just know if their spell succeeded or failed, and beyond that must make some improbable statistics-based guess based on their expected success rate.
Did you fail because your Focus on Men didn't apply to Jacque, who is actually a woman passing as a man? Or is Jacque actually a Faerie and has Might, or a Papal nuncio, or carrying a Relic, or a hedge wizard with some weird not-quite-Resistance that the storyguide doesn't quite remember the details of?
It seems like, while they wouldn't have the numerical specificity of it that we do, they must know about as much as their player about their actual capabilities and how they're being affected by circumstances, because there's no word count given to maintaining separation between those things. It's an expected part of the game that if you, the player, notice your casting totals are being reduced, your magus will notice enough to be cautious about trying their strongest spells, even if they haven't used other senses to deduce the cause. Whatever a given spell might be doing, and whatever having a Focus "feels" like (just being especially talented when you work with magic in a certain way or whatnot), you'd probably have a sense of "Huh, casting that spell didn't go how I expected, it's usually a lot easier than that" or something.
I tend to take it as "you can tell the difference between casting without fatigue, casting with fatigue or failing to cast. You can't innately tell whether a spell has penetrated or not (although it may be obvious in some cases)."
I am very much of the opinion it is a vague idea due to the tone of Ars Majica.
Other games can have the characters being very specific. For example, some games can have a magi thrown a huge blast of fire to incinerate an enemy, and not even singe the person who was in melee combat with the foe. I've played those games a lot, and like them, however, AM is definitely not that. It does not give the players perfect knowledge of distances, chance of outcome, etc.
For example
SG: Voice range is roughly 50 paces.
Player: Whose pace?
SG: A regular type of pace.
P: Can you give me feet, or inches, or metres?
SG: Not really.
P: If the person takes a step back, does the spell not reach them?
SG: Maybe.
Nearly everything should have a rough understanding. For example -
Even if there is an attempt at an empirical experiment, with someone repeating a voice range spell 10 thousand times, with the target moving back and forth, to try to work out the range, it will be inconclusive. The target will take a step back it fails, another step back, it succeeds.
Magic, while it can be studied, it does not obey rules like Newtonian physics does. Due to it's chaotic nature, the mage is never truly sure of chance of success, distances, any variable that is reasonable to be unknown.
I've realised this is a complete thread derailment. Sorry Cardrew.
To answer the OP. I would consider the casting total is originally calculated without using the focus, and must succeed, fail, or succeed with a stamina cost, based on that casting total. If a target affected by the focus interacts with the spell, the bonus of the focus is applied.