Grabbing something from one system and sticking it into another can be a sort of complicated issue.
Generally the consensus is "translate the fluff" and skip the mechanics. In terms of your phrasing, I think it'd be "logical" although your apparent meanings for "logical" and "exact" aren't typically the way that folks think about things when bringing them across.
One reason usually given for not doing an exact conversion is because there can be lots of little subsystems imbedded in something. When you go for an "exact" it will often require either trying to recreate those systems, or trying to remap them onto something that already exists in the new system. As more and more of these things pop up, it gets messier and messier; you get a cascade effect going.
One reason that almost nobody ever brings up and is a fundamental underpinning of the mechanics, is that of "system assumption" as I call it.
See, there's a whole debate about whether or not "system matters" when it comes to rpgs. Some folks say it's important (I'm one of 'em) and others say it doesn't matter. My main argument is because of System Assumption. The rules tend to focus and reward certain things. For example, D&D is about wholesale murder and looting. Most of the rules in the book are about physical combat, the rewards are focused on killing stuff, and the loot is focused on improving combat as well. Roleplaying conflicts is pretty much hand-waved and a single roll or two determines the outcome of the encounter; combat on the other hand takes hours, has multiple rolls, multiple strategies, etc.
I'm not slamming D&D here, just pointing out my perspective so you can understand some of the assumptions I have and will be able to better decide how useful my perspective is for you.
Anyway, System Assumptions...
So, there's all sorts of stuff that a game focuses on and stuff that gets ignored. In some cases, the designer will actually sit down and crunch some math, trying to arrive at numbers that work for what they want or feel the game should be like. So systems that are more "gritty" or "realistic" for example will tend to have lower chances of success, smaller increases in capability, etc.
Do you know the probabilties/assumptions of success for Heavy Gear? How about d20? Lots of folks generally don't know the answer to this question for _either_ of the systems they're messing around with. Even if they do, often it's a hassle to figure out an accurate "translation" from one system to another.
And even if you've got that under your belt, there's the question of fun versus system fidelity. In other words, even if you "accurately" convert a Heavy Gear into d20, is that conversion actually going to be fun to play with? For example, guns in D&D. If you go with an "accurate" translation of guns into D&D, you wind up with weapons that are "overpowered", must have, and generally unfun to be on the receiving end. On the other hand, when guns are put in that are more in harmony with the various assumptions of the system (miniature-based combat, mostly short range/close combat) then you get people talking about how "unrealistic" or "unsatisfying" guns in D&D are.
It's critical you have a clear idea of what the game is trying to achieve and how your conversion fits into it. And recognise that chances are something is going to have to give, because System Assumptions and fun can be at odds with each other in general, as well as within each specific system.
As far as some specific advice?
Get the Mecha Compendium Deluxe from DreamPod. It's a righteous mecha book done by the folks behind Heavy Gear and Jovian Chronicles. They rely on a different version of Mecha rules than the ones put out by WotC; in my opinion, it's a much better version than WotC's. The Deluxe version contains some additional setting and rules material. Since it's explicitly focused on mecha, it'll cover a fair amount of what you're looking to do; it's point-based, so that can give you another tool for trying to limit/control things the way you want.
