Soloing encourages the most time-efficient way to level, which is to grind "green" monsters at a slightly lower level than yourself.

When playing in a group, the leveling rate is slowed by:

1) the least-skilled member of the party
2) higher-level characters reducing the amount of exp you get per kill

I think in theory if you have a party filled with equal-level and equal-skill characters, they could level slightly faster than any of the individual solo characters. But how often does that happen? Actually one way in which it happens is to have a complimentary 2-person party. For example, my paladin (high hit point, low damage) loves grouping with spellcasters or rogues (high damage, low hit point). We grind monsters at more than double my solo rate.

Just curious what is the lvl distance between to players to gain decent exp example, one player is 28 and one is 21 when will it be benifical for them to group and both get decent exp? Like EQ has 7 lvl diffrence

IMO, the biggest factor affecting leveling time when grouping is downtime that grouping incurs. Most people, especially high-level, don't group except to go to instances. You can't find a group for most instances unless you are in a very big guild (100+ active members) or in your capital (e.g. Ironforge for Alliance). So, say, you want to go to Sunken Temple at level 50. First you spend 10-20 minutes in IF yelling "50 warr LFG ST". Another 20 min trying to fill open slots. 10-15 min to travel from IF to the temple. Someone gets lost, someone does not have a flightpath in Blasted Lands, someone has to be summoned. All this time you're getting zero experience. That's a whole hour wasted just to get the group going for a 2-hour instance run. When the instance is over, everyone takes a portal back to Ironforge, and again you are 10-15 min away from the nearest 50+ zone.

Also, experience per hour while actually fighting in a group is normally lower than maximum XP achievable soloing, but that is a different story.

I think alot of this study doesn't account for the fact that MOST people only group up for instances, or to kill one hard mob that can only be done by having 3 oe 4 people.

This risk vs. exp factor is too high on the risk side that I might actually get a bad groupie. So I only group with guild members (rarely), and when I group with people that are outside of my guild, it is usually only for a instance.

I think the big jump in time advantage at higher levels is raiding for equipment. Solo-only characters do not spend time raiding. Grouping characters start raiding for end-game gear more and more as they approach 60. I hypothesize that a significant portion of the time difference at level 55+ is caused by grouping characters NOT working on leveling (gaining XP) but working on raiding instead (gaining equipment).

I'm curious how you correlated this on server population age. On my server, one of the older ones, established players who have no meaningful way to progress outside of raid hours (why spend 100 hours trying to get a modest item upgrade that you can vastly supercede in 20 hours of raiding?), but enjoy the actual gameplay of World of Warcraft (or so is my premise for the consequent:), and so level up alternate characters.

These alternate characters, as activities for off hours, churn rested XP, efficency, and twink factors (that is, they maximize game mechanic advantages provided by their superior experience of having already "played the game", and the difficulty of getting equipment is minimized - how much easier is it to kill that level 25 monster when you're decked out in rare items that a "real" level 23 player would most likely have one of, or have spent hours acquiring that the patron character only need minutes to raise funds for?).

My first character took 22 days of play time (~500 hours) to reach level 60. My second took 10 (~240) - and I, for amusement's sake, handicapped his leveling from 57-60 (so that number would more validly be 7). And that was a solo unfriendly character (in both class, and talent pick).

This independent variable - "Am I not the player's first character?" should wildly skew this data in all kinds of ways (and most probably, in favor of solo hourage).