most weapons crits could be toned down (e.g. only fire on a 2+/3+, -1AD, half AD on your longest ranged weapon, etc.)
To be honest, the simplest solution for weapons crits is to put more granularity into the ship's weapons, as well as the table.
For example; imagine if the omega's forward armament consisted of two seperate heavy lasers of 3AD each, whilst the Olympus gunship only had one such weapon with about the same AD as one of the Omega's guns.
so the Omega would have a statline:
Heavy Laser B 30 3AD Beam, Double Damage
Heavy Laser B 30 3AD Beam, Double Damage
Then you have the ability to put in criticals like 'destroy one weapons system on the nearest facing' or 'destroy D6 weapons systems on the nearest facing' - at which point a big ship does retain the durability of its weapons as well as its hull - since the port broadside of a pulse omega consists of 1,452,673 heavy pulse cannons you shouldn't be able to take them all out with one shot from a delta-v's gun.
From starboard.
It's a bit reminiscent of B5Wars but (provided you don't go mad with the granularity) it adds a bit more realism without much more bookkeeping.
It also adds a way to make certain ship's weapons tougher - an omega, as noted, has two heavy lasers on the prow but a marathon might (for a random example) only have one heavy neutron cannon despite it representing roughly the same quantity of firepower - it just hits like the proverbial tonne of bricks. Loose it, though, and you got nothing, whilst the omega can at least keep shooting on half firepower if one of the lasers is blown off.
It also allows for ships that would otherwise be at the same PL to beseperated by points if one is slightly (in some cases clearly) the better of the two.
Priority levels I like. As noted a points system would let you tweak costs to acknowledge that an early years nova is worth 112% of a third age (or whatever) but the thing is that getting thirty-odd points spare in a thousand-point game doesn't really help unless you have the ability to spend them in a helpful manner - so unless (as noted) you get access to a load of cheap 'ship refits' or 'veteran captain' abilities you're still left with a not-quite-balanced fleet.
Other games with points systems still consistantly fail to produce balanced army/fleet lists. Personally, I think the system is simple, quick to use, and user-friendly. It's good, it's unique, and whilst units aren't always balanced point-for-point, as long as they're not grossly unbalanced it doesn't matter.
Tweaking the grossly unbalanced ships is important but it's done via statline rather than points cost.
I can see an argument for some composition restrictions (to prevent, for example, an all sagitarius fleet when it's theoretically a support ship to start with), but ultimately anything more than 'rare' and 'common' like war of the ring (cannot have as many rare as common) threatens to 'complicate up' fleet selection to the point that you loose the simplicity.
Also, with a lot of fleets, there are only one or two choices at a given priority level anyway, and the nature of splitting down points means you must have a 'core' fighting ship at each priority level or end up sooner or later with a fleet list that forces you to take a million scouts, or troop transports, or whatever (note - I know shadows will do this anyway but the shadow scout is a nasty warship in its own right with the
scout trait being pretty much a bonus).
I do agree, however, that it causes problems that being 'better' is only recognised in multiples of two - whilst I like priority levels, having a ship's cost only increase in a doubling series, not a linear fashion, does mean that, as noted, you regularly feel a ship should be "...and a half". A refitted and improved variant, for example, won't be twice as good - so shouldn't go up a priority level - so shouldn't be better at all. Which feels daft.
A system based on priority but with a few more levels, and where the priority levels were of 'value' 1,2,3,4,5,6,7,8 seems better for that sort of thing than one where they are of 'value' 1,2,4,8,16,32.
However the problem is that you rapidly run into the 'top end' of the scale; prior to Armageddon there was a common complaint that the Shadow Warship (for example) didn't feel scary enough; being able to have the odd ship that is genuinely worth 32 sunhawks helps make different fleets feel VERY different.
However again, a doubling series comes with its own problems - if power doubles at every level, then the 'allowable variance' in power at that level will almost inevitably double - if we dredge up that comment about one variant being worth 112% of the original, then on an armageddon-priority ship that's the equivalent of nearly four 'free' sunhawks of combat power.