Very long design post for light corvette:
The gap between smaller ships like corvettes and star destroyers is predictably huge - after all we are talking about over a thousand fold difference in volume. This makes it difficult to rationalize small ship performance in ways that don't require character shields as such. So when designing this, I had to do some numbers for the frigate and lesser ships.
We have dissipation and reactor numbers from ICS, but not total shield load. For the sake of a starting point, let's say a normal warship can take roughly 10s of its own maximum reactor power dumped into the shield "buffer" before failure of the system, assuming no dissipation (with 1/3 maximum reactor power in dissipation, it'd take 15s to failure). This is assuming 100% hits at maximum power from an identical ship - at a 33% hit rate we're looking at stalemate. 15s of maximum power means 15 full broadsides (1/s) delivered to guaranteed kill (barring local shield panel overloads). The buffer provided by shields and dissipation from shields needs to be long enough for a target ship to react, otherwise ships die really really really damn fast in even duels, which is not something we usually see in SW. But we do know ships are supposed to be fairly dangerous to each other, so this sounds like a reasonable level of relative lethality to shoot for, in a ship not optimized for maximum relative survival; for a ship that IS, take the AOTC Nubian barge that can dissipate 2x its reactor output in shield heat. I'd expect Mon Cal ships (by reputation) to dissipate heat on the order of maximum power, which would give the 1.2km versions durability parity with ISDs dissipating at 35%, but not really weapons parity.
The problem is when we get to scaling for mismatches. Scale so a small ship can too easily survive a bigger one's secondary battery hit (basically increasing total shield load), and it will become totally indestructible against ships its own size. Scale too little and a large ship can kill a small one by accident with its point defense (you *would* expect a big ship to be able to use its main battery to casually annihilate a much smaller ship).
Going by the corvette:LTL :: frigate:MTL :: destroyer+:HTL comparison, from a tactical perspective, it seems like you'd want a normal corvette to roughly be able to take an MTL hit (200GT) and not melt immediately - maybe lose almost all shielding and be kicked sideways like an empty can, but *alive* to try and run away. So, assuming shield load at 10s of reactor power, here are the numbers based on surviving 1 hit of:
Hit Shield load Inferred reactor power Inferred shield dissipation
1x6MT laser cannon 2.5e16J 2.5e15W 8.8e15W
1x500MT LTL 2.1e18J 2.1e17W 7.4e16W
1x200GT MTL 8.4e20J 8.4e19W 2.9e18W
1x40TT HTL 1.7e23J 1.7e22W 6.0e21W
So, a 150m corvette to warship standards should be targeting a ~1e20W reactor. Note a ship able to take 1 40TT HTL hit and keep going is at 1.7e22W inferred reactor power, or a Velox, which I aimed to plausibly generate 2e22W (yay consistency). Scaling down from there, a Vigil should generate something like ...5e21W.
So, next question is how to scale the reactors appropriately, such that the power matches the aims above and is consistent with available data. At the same power density as an ISD, one would expect an ISD-proportioned ship at corvette size (~150m) to be outputting roughly 1e22W. This smells like bullshit, since Acclamators at 700+m are only 20x more than this (yes, Acclamators are probably de-tuned from a full warship, but it is a hard data point). From the lower end, scaling to the AOTC Nubian barge (assuming the front half of each nacelle is reactor, say a 10m x4.5m diameter cylinder), yields roughly 630 cubic meters, or roughly 1/13000th ISD power (1.9e21W, vs actual performance of 3e18W, or 600x more than actual, but it is basically a transport, not a warship). Moving up the scale, Munificents at a very skeletal 825m yields 2e23W - whereas from a pure "fit largest possible reactor in hull" would yield ~5-6e23W as a predicted power point. Of course a Munificent actually has smaller dual reactors compared to a single large reactor so power density may actually be similar to that of an ISD. It may not even be able to mount a big one given that it is such a skeletal design.
It seems like the sub-destroyer designs are generating much less power than hypothetically possible on the hull form. Why? Certainly this could be due to emphasis on something other than power generation, but that seems a bit daft design-wise - warships should maximize power whenever possible. Another possibility is due to poorer efficiency for hypermatter reactors at small volumes.
This suggests If you physically can mount a big reactor, that increases your power-to-weight. All this combines to drive high performance combat ship design towards low surface area high volume configurations, to maximize reactor and minimize the area (and thus increase the relative strength) of shielding.
One estimate I've seen has Tantive taking 30-ish hits from Devastator to shield failure over ~20-30s. I think these ISD guns are not likely to be gigaton scale weapons; no documentary evidence from models or other sources exist for Acclamator-sized guns on an ISD, and the teraton-scale heavy guns did not bear during the chase, generally speaking. If the light guns are 6 megaton lasers from Acclamator, then you have a pretty unimpressive total shield load for the CR90, with dissipation at most unable to stop ~300MT/s (1.2e18W dissipation, which assuming Acclamator performance suggests reactor power on the order of 4e18W). This ship is performing at the same level as a 40m yacht (using the barge numbers would actually depress potential reactor capability on the CR90, as the Nubian ship prioritizes dissipation over total power). Not a good look for what seems to be a fairly ubiquitous small warship, even if one that may be largely civilian-spec.
Let's assume then that the 6MT Acclamator weapons are actually just "laser canon" in warship terms (the ICS doesn't actually call them turbolasers). What would a real LTL look like? HTL -> MTL scaling suggests about a 200-1000 fold difference in weapon classes (assuming an ISDII mounts 40TT guns). Given 200GT MTL as standard, LTL would start picking up just beneath gigaton scale. If we assume 500MT as a normal anti-ship LTL (as opposed to a self-defense laser), then the CR90 starts looking much more attractive - dissipation would rise to 1e20W, and reactor scales now to ~3e20W. Much more respectable, and roughly in line with scaling from the Naboo ship. Obviously there are many fudge factors in this calculation - Tantive could've taken many more hits, ISD light weapons could be gigaton scale, etc.
The CR90 though is essentially a very narrow tube. That dramatically restricts the potential reactor volume for the length. But a value between the two extremes (4e18W - 3e20W) would settle down at just surviving 1 200GT MTL hit, so roughly 1e20W. The new ship has a much fuller hull form for the length, so there should be little problem mounting a much bigger reactor. The new light corvette scaled further should generate less than 1e21W, so hypothetically 5e20W.