Omg guys, looks like you have your own local speech here

Its not that any of you have bad english, its just posts are so long and you keep repeating same things and its hard to follow what exacly are your trying to say

And also a lot of new information
Just for some content - post contains 2 parts, first is about responses in general and second is horses vs longbowman combo calculations
Northen wolf - if you want help to improve i guess i can give some constructive criticism:
1)Try to not repeat stuff other player already said(your spoilers about defence, attack and shields of each unit) and even if you do dont do it sooooo long;D On my resolution those 4 "spoiler" thingies takes 1 full screen.
2)Try to build a plan in your mind, and if very long post, maybe you even want to write down points of plan. This is required to minimize repeating(and remarks) and orginise thoughts as elegant as possible. For example:
I want to say this thing from myself as introduction and short answer to question
reply to some particular post.
Then i will say some in-depth analysis based on facts(and/or my theories)
Then i will try to predict future outcome in possible situations.
With this i want to finish.
Dont get me wrong, you have some of that, much better then a LOT of people, but its hard to figure out where something starts and where finishes. And then again those people dont dare to write so long posts i guess

Kinda feels that everythings blurred together and stuff is repeating.
And also i know what you will say lalala i do that myself, but at least my post wasnt so big, it was only like 1/3 of my screen;D
You know what i just gave myself some pointers lolx

Anyway lets move on.
3) When you say something as an answer(in this case) try to not state only the obvious(that the archers have no 1 defence), but instead give some reasoning with it, i'd rather say something like:
"Archers have 1 defence and therefore they cant do this and that. They also have only 1 movement speed, so they cannot do this and that. So i would always choose horses and MI over archers",
instead of making it long post as a:
"Archers have 1 defence, as opposed to Medieval infantry who have 2 defence. And horses have 2 movement speed as opposed to archers who have only 1. Horses are better because of that movement speed. And Medieval infantry is MUCH better cos of their 2 defence. I dont want to say anything bad but i think archers sux."
Understood what i'm trying to say?
Its not that i try to say that you suck with your post, i am not sure if its true, maybe my point of view is the bad one. What i'm trying to say is that with my point of view its very hard to follow what exacly is your point(s).
So back to the topic

Thanks a lot for info about bridges. There are probably a lot of cases where Knights would be much slower without "Engeneering" tech, since usually theres a LOT of rivers on map. Also i wonder if extra speed works on neutral territory as well as your own after engeneering? I'm guessing that it doesnt have effect on enemy territory...
Upgrading is not always an option, for everyone in every game. So you tend to not upgrade archers as you already know how to make MDI. You don't build longbow all that often.
I do actually

I always use medival infantry only for some insurance. Mostly army consists longbowman or knights, or i just skip that tech completely if i am playing cultural war or space race.
If my enemy is bring knights, I want to have knights, not LB or MDI.
Why not? It feels comfortable with me.
I usually have 2 stacks of Longbowman/Musketman or Pikeman/Trebuchet sieging some strategically critical opponent cities(for purposes of roads, to make other troops get to destination faster, or some city with good wonder in it, theres very rarely a chance of cutting off valuable resource with this strategy).
Then i have some Horses to secure perimeter(usually 2 packs, one at the most far side of my kingdom and one at the place where opponent could retaliate, since longbows lack speed and cant catch counter-attacks)
But i have those in very good(imo) precenteges - about 60% offencive army 15-20% horses and 20-25% defenders. It all also depends on situation of course. Precentages change if i'm surrounded from all sides on Panagea map. Usually have more horses and more defences(well it sux anyway to be in the middle i guess unless you take out some of the sides out early).
Point of all this over knights myself is - i have much more cost-effective army. Slower, but i have few casulties. I have trouble to explain it better, so i'll just give a situation:
"I have 20 Longbows 6 musketman and 12 trebuchet divided into two stacks, attacking two enemy cities simultaniously. Opponent have like 3-4 defenders in each city. I bombard as much as possible, and attack with longbows. Usually casulties are 1-2 longbowman per city."
After taking first city Opponent(AI) will attack with his knights and try to take out my longbows, they are protected by 3 musketman and by the time i take city i add 1 knight for additional 3 defence.
Now the main question is: how many knights will he have???
I estimate around 10-15, since they are really expensive and he isnt as good prepared for war as i am because he didnt know i would attack him, so i should have slight advantage in these terms.
So lets go with say number 12. Usually he have 2 options eather attack with 6 knights one stack and with other 6 other stack or do something like 10-8 knights to one stack and 2-4 knights test luck to other stack, i find it very rarely that AI can mobilise whole army on one stack. Lets go with worst case scenario and say he attacks with 10 knights(of which not all might be veterans but lets say they are).
So he attacks with 10 knights pack of 3 fortified musketman+terrain bonus and a lot of trebuchet - roughly that makes 3 knights against each musketman.
So
1) Each musketman have 1 more hp then knight(or maybe even 2 more hp since horses tend to retreat on last hp) that makes 2 horse hp against 4 musketman hp. That reduces odds from 3/1 to 1.5/1, or if they might not retreat then to 2/1.
2) Fortification gives 50% more defence, so from 4/4 - attack/defence it makes 4/6=musketman are 50% more effective.
3) Terrain bonus, i usually try to be on jungle or forest, if possible on hill. Usually its 25% or 50% more bonus. Now i'm not really sure how that calculates but i'd go with 50% from fortification and 50%*50%=25% more from terrain, or 50%*25%*=12.5%. So rougly it gives something like 0.7-1.2 more defence value.
CONCLUSION:
MATH: 4*3(three knights) offence against 4*1(one musketman) defence + [4*1.75(fortified*hills)] * 2(bombard+retreat)=12 against 14 odds. NOTE: its if theres 9 knights, 10 knights should about even the odds.
IN WORDS: In best case scenario its 50/50 that 3 musketman will take out 10 knights.
Worst case scenario if you have only 10% grassland bonus and maybe stack is split, due to needed musketmen mobilization to newly captured city(and since trebuchet moves very slowly), and there would be left 1 musketman and 1 newly added horse or just random medieval infantry for insurance, odds would go down twice, meaning there would be 25/75 chance that 1 musketman and 1 something will take out 10 knights(until 1 hp). Or it might make something like 10/90 if knights dont run away and fitht till end. So from attack of 10 knights you will usually lose defencive units and maybe 2-3 longbowman.
FINAL CONCLUSION: YOU LOSE VERY FEW UNITS and TAKE OUT MOST OPPONENT ARMY ON HIS GROUNDS, with roads and everything, and also meaning that pressure is on opponent and risk if he loses is bigger then you lose your army. KNIGHTS thumbs down LONGBOWMAN combo

Have you tryed modern armor/radar artilery/mech infatry combo? If opponent is 1 tech behind you, say tanks and infantry, stack of 2 elite modern armor, 7 radar artilery and 3 mech infantry seems pretty immortal
Tryed in one of my recent games, modern armors attack 3 times per turn on 1 hp infantry, from artilery

NO LOSSES WHATSOEVER. TAKE OUT ANYTHING OPPONENT CAN THROW AT YOU MWAHAHAHA
DD