My current thoughts:
Morrigan, Harbinger of Strife: Earn

,

, and

Points from kills (1CS = 1 yield scaling with gamespeed). +2

from Barracks, Armories and Citadels.
Epona, the Great Mare: +10

, +10

, and +10

when your Borders expand, scaling with Era.
Bran, the Sleeping Guardian: +10 HP Healed in friendly territory, +25% Growth, and +12

when a Citizen is born, scaling with era.
Lugh, the Skilled One: +3

, +3

, and +3

in Cities with a Specialist
, scaling with era. or No change.
Dagda, the All-Father: +1

, +1

, +1

, and +1

for every 4 Followers of your Pantheon in owned cities. (No change)
Ogma, the Learned: +1

for every 3 Citizens in a city, and +3

+3

Great Scientist points in the

Capital. +

equal to science per turn when you discover a tech.
Nuada, the Silver-Handed King: +1

for every 10

per turn, and +1 Golden Age

Points for every 5

per turn. +2

Gold per Market and City Connection.
Rhiannon, Goddess of Sovereignty: +2

, +2

+2

, and +2

in every City. No

from isolation.
Mannanan, Son of the Sea: +3

, +3

, and +3

in coastal Cities.+1

and

from fishing boats and Lighthouses.
Cernunnos, the Horned Stag: +1

and

from Forests, +1

and

from Jungles. +1

from Camps and + 1

,

&

from plantations.
Okay the reasoning:
Morrigan needs the production bonuses or it's simply worse than the current Lugh until late game. Just 5 cities working markets is 20



per turn. Are you killing 20 CS of units per turn? I don't think so until much later, especially because you'll probably have more than 5 cities. The yields should also scale with game speed. This is obviously for hyper-aggressive or surrounded authority players.
--
Epona is obviously perfect for border-blob, and post-nerf hopefully manageable at that.
--
Bran should be a defensive pantheon, allowing you to survive as tradition against aggro enemies. I think +10 HP healed is more valuable than stacking city attack bonuses, and the culture on birth is best here. The

might be high, but I'd rather nerf it later if needed because buffing is agonizing.
--
Lugh seems too front-loaded. It's too good early and bad late, I feel like snowball or die is a bad design. It seems suited for progress.
--
Dagda
could be nerfed like this: "Dagda, the All-Father: +1

, +1

, +1

, and +1

for every
4 Followers of your Pantheon in a city." Local yields means having 2 followers in 2 cities each won't give you any yields, as opposed to currently where you get 1 in your capital. Also followers over citizens mean cities will normally need to be size 5 or 6 to get their first bonus.
HOWEVER You're really over-valuing it. (As was I.) You need 16 followers per city for it to outpace the current Lugh aside from production. If we're talking raw yields you need 12 followers per city to match. That takes
quite a while, while Lugh comes online at trade. This seems suited towards tradition, especially when you think you
might have a lot of room to expand.
--
Ogma is the science victory belief. It's meant for when you're tradition and you've got your plan.
--
Nuada seems good for wide play. The early gold leads to late culture unmatched by anything.
--
Rhiannon could be re-worked. I tried to offer it a niche by allowing you to be mini-indonesia, but it seems insanely niche.
--
Manana Wet banana needs this buff imo. I've tried it 2 or 3 times and it feels like the wrong choice every time. It also seems like the wrong-choice in hindsight. Obviously this one is location-based.
--
Cernunnos, the Horned Stag: Another location-based pantheon. I think we agree on it.
So that's my thoughts. I can see good reasons to pick every pantheon I listed, and I don't think any are overlapping in a serious way. The worst seems to be Lugh and Dagda, but I think I've given them both separate uses as explained.
Thoughts?