What perfects did was effectively cap the power level, or at least they were supposed to. You could never get a better defense than Heavenly Guardian Defense/Seven Shadow Evasion, and no attack could pierce them. The idea, I believe, was to force the scale to stop growing. Basically once you got to ‘parry landslide’ there was no point statting up attacks for anything more dangerous, because the optimal solution was always HGD.
Since Solars were also the strongest Exalts, and thus strongest playable type, by default this also was supposed to effectively cap the entire game line. “Look guys, there is no point in adding attacks for levelling mountain ranges as anything but story effects since HGD just works.” The point was to game outward, rather than upward. To work in that range between mortals and Solars as the established levels of power.
Of course, many writers didn’t want to do this and we ended up with the scenario you describe. Eventually someone wanted to power scale upwards, and so we ended up with transperfect combat; that is not combat which can overcome perfects, but combat built entirely around their existence as high-level building blocks. This started in middle 1e with Sidereals, which contained our first high level combat charms (we also see here the first use of perfects as bread-and-butter effects, with Impeding the Flow being the first 3m semi-perfect, i.e. it always worked against anything it was possible for you to parry). When 2e came around they tapped Borgstrom, the author of 1e Sidereals, as the author of the charm chapters. It is thus not surprising that she basically took the same attitude to 2e combat that she did to 1e combat, with perfects being a specific utility effect of the basic combat system and balancing everything around them.
This was probably a mistake. It compressed what should have been the high end of the game down even further, leaving less available conflict in the game between mortal and “well everyone is perfecting everything now so we’re not even rolling dice at this point” than 1e had. This meant when the Ink Monkeys eventually got involved, they jumped straight into transperfect combat which led to all sorts of cosmic-scale bullshit.
The game should be designed such that you establish a maximum power level and perfects create a ceiling for that. Once you hit the perfect range, the game ceases to be the game. Then perfects should force it back down. You can’t overcome perfects with spammed perfect-or-die effects, or you shouldn’t be able to, anyway. You should beat it by attacking in the range of effects that is just below the perfect range so the game is forced down the the maximum scale you find utility in representing in play.
(in response to someone asking if that meant perfects should be Essence 5 capstone charms)
Hmm. I’d say the opposite. Instead of having them be Essence 5 top tier effects, have them be incredibly low Essence (say, 2) effects available in chargen. The trick would be they should be prohibitively expensive to apply against any effect which isn’t an insta-kill. Perfect effects are panic buttons and should only apply to effects outside of the normal range of combat.
Obviously this requires balancing combat such that peer level combatants aren’t spamming insta-kill level effects. This isn’t just in the literal sense of save or die (or no save as the case may be) effects, but that damage scaling has to be survivable as well. So if a standard two dot artifact weapon is pumping damage up to 10d post soak and you only have 7 HL, then you aren’t balancing combat effectively on the peer scale and forcing the perfect into effect more than it should be.
This requires a rigorous game built from the ground up to accommodate a certain scale of action; say anything from 1d to 20d is ‘acceptable’ damage and 21d+ is the realm of Perfect Defenses. You need to make certain that no combination of attributes, abilities, mutations, equipment/artifacts and charms/sorcery can easily pump damage up to 21+d, at least not without being significantly more expensive than the cost of a perfect defense.
For example, if PD costs 10m, 1wp per use than the minimum cost for getting 21+d of damage should be 20m, 2wp. This includes multiple actions (if you attack twice for 11d damage, then you damage is 22d and should be costed appropriately). Though better to just eliminate multiple attacks entirely and leave that to stunting.
This requires you to have a more robust set of combat effects than “I increase damage” or “I hit multiple times” to keep the game interesting, but I don’t think I have a problem with that. High end Charms would then be less about spamming ultimate death ray attacks and more about mastery of the sub 20d damage environment.