I also don't like the Allies. The speed penalty is felt in ALL kinds of battles: land rushes, naval encounters, tactical manoever battles, the list goes on. The research buff might be, as you suggest, highly questionable. But is that really true? Let's examine.
As I see it and as I use research, I have a number of "key units". They are usually one or two ranged artilleries, a rush unit, infantry or militia, sometimes a medium tank or TD, three or four air units, and two or three naval units. Lets say 10 total; these I keep up to date and research new levels as soon as I can. Then there's a number of support units: AA, AT, commandos, strats, something secret. I typically research one or two levels of the ones I pick in a round; either way that's small stuff, the bulk of research time goes to the key units.
With non-allied tech, you can just barely keep up researching the key units as the new levels become available, and a few of them are slipping behind a bit. That's not too serious, you can pick which one of them you let slipping as the game progresses and decide based on what kind of battles you are fighting in every stage, and foresee in the near future. That means that in a battle where you are actively using seven of your key units, maybe two of them are a level "behind" (plus the old stuff you are still using of course; more on that in a bit). New levels are about 25-35% more effective than their previous version; so you are about (2/7) * 30% = 8.5 % less effective than optimal level. That's not completely fair because a good leader lets the LEAST important of those seven slip; so lets make a rough estimate of that and say it is actually only 7%.
Lets say for a minute that Allied leaders do NOT have to let key units slip behind, and they can keep all those 10 key units up to date all the time. So basically, they don't suffer that 7% tech penalty. That's less than the Axis power buff. But it is not just about power: stuff like HP, speed, flight range, and stealth are affected as well. On the other hand, you need to invest more in industry to produce the same amount of units (though this is again offset by the third allied buff, 30% less build time). All in all, I'd say it was pretty close in value to the axis buff, maybe slightly less.
Then there's the upgrading buff. As discussed in a different thread, I don't really like upgrading and prefer building new units; on the other hand, when you can get TWO levels it becomes a different story, and there are some other reasons as well. I don't have hard data for this, but as a rough estimate, I think I'm upgrading something (anything) like 30 times in a round (100p map). Since upgrading costs "half a unit" I spend 15 units' cost on that. If an allied leader does the same, he spends 20% less on that. So the value of this buff is 20% * 15 units = 3 units over the course of a round. It is something, but it isn't VERY much.
So all in all, it is a pretty decent buff. It is very unfortunate that it is coupled with the speed penalty, which I think is the worst debuff of all; imho, it makes the allies almost unplayable (and yes, pan-asian my favorite). Since allies also don't have a buffed rush unit (mech inf doesn't count!), it will remain my least favored doctrine.
