i don't know what your fascination with relative gains is we aren't looking at that we are looking at end results
Relative gains are what's important when balancing a support unit as that tells you what a support unit actually adds to the things he support. Absolute gains are irrelevant as those are completly dependent on context if they're weak or powerfull, whereas relative gains are context independent. An absolute gain of 10 can be asthonishingly good in one situation, and barely noticeable in another. A relative gain of a 100% is going to be noticeable in any situation.
Simply put, if the relative gains are too insignificant you'l never care about the hero. Your opponent needs to want to kill the support hero given the chance because he's a significant power multiplier. But if his relative gains are too little it's just gonna be easier and more efficient to kill whatever he's supporting and ignore the support hero until he no longer has anything left to support. To illustrate you want to kill our skink priests cuz they have powerfull support making whatever he supports significantly faster and sturdier, you want to kill an oldblood cuz you've already decimated the 40 man saurus horde he was supporting, and compared to that last handfull of strugglers he's a bigger threat...
Now of course, once you've determined the relative gain to be balanced you're going to have to look if there's any specific outliers in which the absolute values become too much. And how to deal with these outliers. But that's only going to be your second step when it concerns a support unit as the support he gives should, first and foremost, be significant. And re-rolls of 1 in an awkwardly sized little bublle on a slow hero are just too insignificant to become a "support" hero.
Compare it to other support heroes:
- Skink priest: reroll all saves, charges & runs, either on 1 target, or in an 18" bubble. Also prayers once we finally get those.
- Skink starpriest: double damage on a 6+, various spells, potentially prayers.
- SCE lord castelant: +1 to saves, saves heal on a 7+.
- Auric runemaster: a permanent debuff and has acces to prayers.
- Necromancer: heals 2 nearby units, can cause a unit to attack twice.
Those are proper support heroes. Their buffs have significant relative gains, or are permanent in the case of the runemaster's debuff. They have acces to multiple effects, that potentially synergize. And are able to affect multiple targets in a single turn. Within a far greater range. And before you point out that the oldblood is a better combatant than the skinks & necromancer, the runemaster and especially the castelant are capable combatants themselves. This is just looking at their warscrolls, and ignoring that for example the runemaster also procs the hearthguard's berserkers "duty unto death rule" or any potential other supportive abilities they get from battalions, artifacts etc (the SCE have an artifact or trait that give a bubble of +2 to saves for allies if he stands still in a turn, which on a castelant leads to a hilarious +3 healing on a 7+ save).
Do you see the difference between these support heroes and the oldblood's attempt at "support"?
As for it potentially being overpowered if you make it better. That's why you can put in limitations. I never said the oldblood needs to give a horde of 40 warriors mortal wounds on a 6+, that'd probably be problematic. Though based on the stories
@LizardWizard occasionally tells about tournaments, it probably wouldn't even be that much of an outlier.
As for it being too much combined with all our other support. Again, limitations are your friend. And perhaps those other sources should also be looked at. Maybe take away the sunblood's command ability and make him double down on combat, while the oldblood focusses on supporting and combines his aura with the sunblood's command ability in some way. But calling the oldblood "support" and justifying an underwhelming melee profile on account of a fairly weak and awkwardly sized buff isn't the anwser at least.
ok im a bit burned out with heroes lets fix a unit
Kroxigor 200 for 3
Move 8"
bravery 10
wounds 4
save 4+
a kroxigor can be armed with either a drake maul or a moon hammer
Drake bight maul 4 2" 4/3/-2/2
moon Hammer 4 2" 3/3/-/2
Vise like jaws 1 1 3/*/*/*
abilities
Energy transference unchanged
sweeping blows: when a Kroxigor attacks with a moon hammer raise it's attack stat by 1 for every 10 models in the target unit
vise like jaws: this attack does not make a wound role if it hits deal d3 mortals
Bit on the squishy side, quite expensive & the lack of rend means the moon hammer needs quite a lot of targets to pull even with the maul. Maybe add some rend or have it gain attacks per 5 models?
Also, I'd change energy transgerence. It's not bad (in this context..) but the range is quite short, especially considering skinks aren't frontline fighters while kroxigor are. Unless the intend is to have them escort stegadons & troglodons it's just a bit of an odd rule.