The following is based on a no cp pvp environement:
Sharpened:
Useless vs. shields. Useless vs. targets you overpenetrate (very unlikely except for niche builds). Does not affect healing.
4,2% penetration – If you take target to exactly 0% resistance, this is a 4,4% damage increase. If target has exactly cap resist (50%), this is a 8,4% damage increase. Most targets will of course be somewhere in between, and most probably closer to 5k-15k armor/resists after various debuffs.
5k armor/resist target (after debuffs), sharpened will give a 4,5% damage increase
10k armor/resist target (after debuffs), sharpened will give a 5% damage increase.
20k armor/resist target (after debuffs), sharpened will give a 6% damage increase.
30k armor/resist target (after debuffs), sharpened will give a 7,7% damage increase.
Most targets will probably be in the 5k-20k range after your debuffs, so generally a 4,5%-6% damage increase
Precise:
Useless vs. shields. Useless vs. 3300+ crit resist targets. Does not increase proc based dmg. Potential for more burst with lucky crits.
0 crit resist target and 50% crit dmg modifier. Precise will give 4,5% dmg and 4,5% healing
27,4% crit resist target and 50% crit modifier (7 impen armor). Precise will give 2% dmg and 4,5% healing (typical pvp scenario)
27,4% crit resist target and 62% crit modifier (7 impen armor and shadow mundus). Precise will give 3,1% dmg and 5,6% healing.
Nightblades will gain slightly more value due to the 10% increased crit dmg.
Nirnhoned:
Affects both healing and damage. Works vs. all targets. Does not affect proc based dmg.
Nirhoned is a fixed damage increase, only modified by your +% weapon/spell damage increases, normally major brutality/spellpower.
This makes it hard to compare to % dmg increases like precise and sharpened, since the % increase from it depends on the specific skills you use, and your overall stats.
You basically need to see how much weapon dmg/spell dmg your weapon of choise will give you, add % modifiers, and then calculate the value gained based on the skills weapon/spell dmg coefficient. Compare the increase to the base dmg of skill, and you have the % increase.
For exaple on a magicka sorc with 3 sorcerer skills and major spell power buff, using a gold,staff, your crystal fragments damage will be increased by 406 value. (1335*0,15*1,26*1,61=406).
9000 crystal frags tooltip, nirnhoned will increase the crystal fragments by 406 dmg. This equals 4,5%
10500 crystal frags tooltip, nirnhoned will increase the crystal fragments by 406 dmg. This equals 3,9%
12000 crystal frags tooltip, nirnhoned will increase the crystal fragments by 406 dmg. This equals 3,4%
Another example, nightblade with major brutality and using a gold 2hander.
5000 surprise attack tooltip, nirnhoned will increase the surprise attack by 305 dmg. This equals 6,1%.
6000 surprise attack tooltip, nirnhoned will increase the surprise attack by 305 dmg. This equals 5%.
7000 surprise attack tooltip, nirnhoned will increase the surprise attack by 305 dmg. This equals 4,4%.
Generally nirnhoned will be best if you have low damage modifier from stamina/weapon dmg or magicka/spell dmg, and worse if you already have high damage modifiers.
It also scales slightly better with 2handed swords/axes/mauls because of the higher base damage, and with the mainhand weapon in a dual wield setup (since you gain almost full effect from just nirnhoned on the mainhand weapon)
Infused:
Even harder to calculate.
You will have to do a rotation on a player with your setup (but with a trait that doesn't affect damage, for example no trait/training) This alone is probably pretty hard to do, who wants to gold out a non trait weapon?
Do your normal damage rotation on that player, and use a damage meter addon to see your dps.
Afterwards do the same rotation but with the infused weapon and glyph of choise (probably oblivion)
To get an accurate result, you probably need to do quite a few tests and average them.
Bear in mind the higher armor the player has, the more effective the oblivion glyph will be if you use that one. A test against an average armor value target will probably be best.
Alternatively run recount addon while using a non infused weapon with an oblivion glyph for an evening of pvp. Don't fight any mobs or guards. Check how large a % of your total dmg is oblivion enchant. Multiply by 2 and 1,24 due to half cooldown from infused trait and 6% nerfed dmg, but 30% buff from infused on pts. See how much this will increase your total dmg.
Conclusion:
For damage, precise will be the weakest of the 4 traits, even when using shadow mundus. However it does provide good healing, especially if using shadow mundus.
Nirnhoned seems pretty good, it works against all targets and buffs both damage and healing. However if you have high offensive stats it looses effectiveness.
Sharpened outperforms other traits against high armor/spell resist targets by a pretty large margin. It also increases your proc based dmg. But the disadvantage is it provides no healing.
Infused is abit of a special case. It is mainly useful for single target builds, due to how glyphs function, provides no healing, but will in most cases give the highest single target damage by far with an oblivion damage glyph.
I have not done the test above for my own build, but my addon shows that oblivion damage glyph on live accounts for 5-9% of my total damage on a non infused weapon.
Lets say 7% on average. Reduce cooldown by 50%, and increase damage of the proc by 24%, numbers will be around 17% of my overall damage. (which means around a 10% increase of my total dmg from the infused trait).
Now my build does not have high weapon damage or stamina, so for some builds the damage increase will be lower, but still significant.
But it will beat all the other traits by a pretty large margin regardless, in any single target encounter.
If you have any questions to the math used to determine % bonuses, ask and I will explain as good as I can.
And let me know if I made any calculation errors