Seals and spacers
The stock seals had two lips plus four rubber teeth in the front. The two lips were probably supposed to last more miles while the rubber teeth serve as built-in spacers to keep some distance from the bearing (which is necessary because in the RG the seal is on the "wrong" side of the bearing):

These old seals are difficult to find, plus the two lips put a limit on the rotational speed they can do, plus they were too thick for the new cranks. The new design uses standard high speed seals easier to find (high speed = single lip), plus custom spacers made in avional (an aluminium light alloy):

In case you are concerned about the reduced clearance between seal and bearing, do not, as the absence of the second lip compensates for it in excess (that is, the seal should be even better lubed than before though the assebly is more compact).

Bearings
They didn't want me to purchase the original replacement parts as They use bearings with different specs, plus the new crank design uses one extra kind of bearing (three kinds total, instead of two) that had to be self made in any case.
So we took a dozen of Their bearings, disassembled them and had them machined. Then, reassembled everything. Here you can see the three different kind of bearings: the pinned-grooved (known as NP), the just pinned (known as P), and the new one, which I named "P05":

P05 has the pin in a different position (necessary to make the new crank compatible with the stock crankcase). It was discussed for months; I took every possible care in the design to ensure the bearing is as safe a the stock one and the pin mounting is even more stable. Actual machining was done by a very skilled person with extremely expensive and precise tools.

Above you can see the "central group" with bearings, spacers and seals in place: isn't it nice? ;-)

Pins
Even the bearing pins were discussed for months. We considered different solutions and we purchased several different types until, just before reassembling the bearings, we found pins that are exactly the same as stock RG ones (sort of a torsional spring, that is, a rolled sheet of spring steel).
Pins we used in bearings type "P" and "NP" are exactly same diameter and length as stock RG ones, while model "P05" uses both a different diameter and a different length.

Rods
I believe the pictures speak for themselves. These rods are the result of years of testing in the most severe racing conditions and can be abused more than you could ever dream. Their normal usage is at 14-15000 RPM with tons of HP on them, but I know people often over rev them a lot with no damage.

Here you also see the silver-plated shims with dimples for oil retention (of course there are also nice roller bearings, but I didn't take a picture). Then, the rod seats in the cranks, the clearances, the big pin tolerance... everything up to the piston is exactly either as per Their specification or Their standard parts. No need to reinvent the wheel.

And remember, these rods were born for a 54.4 mm stroke. In case you'd want them on a 50.6 mm stroke crank (less solicitations due to alternate forces) you could run them even faster!

Alignment reference
The standard method of aligning the RG cranks was like in the left picture. According to the gammalist knowledge, there was no special tool for doing it, and the result relied on the operator art skill.
I was a bit scared by this as I'm not an artist and could not do it precisely that way.

The new cranks have a precise reference for alignment that, if used in combination with a straight and sharp point, allow everybody to get the cranks aligned with a minimal error. The alignment reference (call it a "scratch") is now made at the opposite side (180 degrees) than the rod as the rod area is thin and highly solicited and must be more than perfect. Even a scratch in that area could compromise its safety and lead to a fatigue stress crack in the long term. The scratch made at 180 degrees is perfectly safe (there is nothing there) and the mounting procedure is the same except you have to count the gear teeth in the opposite direction.

Ignition references
I had the pick-up notches built in two variants: stock 26-11 degrees and the modified 30-18 degrees you see in the picture. The idea being it is to get a more precise advance where you need it, provided you have a mappable (programmable) ignition. I'll post an explanation of this as soon as I have time to write it.

Of course, if you want pick-up notches different than 26-11 or 30-18 it's a relatively minor problem, though for cost reasons it would still be better to make the cranks all the same.

Intake flow
The new cranks shoulder has a shape more favourable to the intake air flow. There are no more edges and jumps and it blends to the case intake port nicely. I spent a few sleepless nights to redesign the intake port keeping a balance between contrasting requirements (maximizing the section area, having it as straight as possible, having a curvature with the derivatives under control, taking away as less material as possible from the cases, not repositioning the carburettors too low, making the cranks as safe as possible in the rod area...)

The new shape corresponds to the last millimeters of the intake I've redesigned; I mean to provide templates for working the intake port accordingly, but I am optimistic this shape will in any case work better than stock cranks' one even if you leave your intake port as it is now.

Balancing
The new cranks have inserts in different materials to accomplish balancing (the round objects you see press fitted into the big holes in the picture). Balancing can be customized as needed by replacing the inserts with different ones (different materials, empty inside, etc. etc.). I don't think anybody has ever considered rebalancing the RG cranks when, for example, adopting oversized pistons, however that's what you should do.

A 54.4 mm stroke crank running at 15000 rpm generates, due to the unbalanced masses, about twice and a half the alternate forces on the bearings and the crankcase than a 50.6 mm crank at 10000 RPM does. An accurate balancing is critical for the safety of the whole. Balancing of 2T cranks is usually only done statically, as they are normally symmetrical and dynamic balancing is intrinsic. As the RG cranks are not symmetrical (I posted a long article about this on the gammalist, time ago), I had them also balanced dynamically that is, we put more weight on the thinner flywheel and vice versa in order to compensate for it and reduce (ideally, eliminate) the nasty side vibrations it could produce.

Crankcase free volume
The balancing holes also allow to tune crankcase compression ratio, depending on what you put in them. In any case the crankcase compression ratio is increased due to the higher volume of the crank parts. I do not want to discuss what compression ratio is better, I just say that cloning Their bottom end as closely as possible is what will work better together with Their top end. There are also other fluid dynamic effects taken into account in the crank design that I can't tell, but I am optimist they should provide better performances even in the 50.6 mm stroke / stock cylinders combination.

Weight
For those concerned about weight, these cranks probably weight more than stock ones (I didn't check, yet) but also weight less than Theirs.
More precisely, they weight about half a pound less each (can't tell more precisely), that is, a couple of pounds less the whole assembly. In a racing world where one gram makes a difference, this is an enormous difference! It wasn't my choice, this is an unavoidable consequence of fitting Their crank design into the RG crankcases (that are smaller). I tried to make the cranks as big as possible but still, they are small.
I know some of you don't care about them being already two pounds lighter than Theirs, and are actually disappointed as you wished them to be even lighter than stock RG ones... then try to think about it like this: call them "reinforced" cranks. Sometimes, like e.g. oversized brake rotors, you need something bigger/heavier for safety (if it wasn't like so, why do you think Hayabusas don't use e.g. moped wheels and a bicycle transmission chain?). Here we are increasing solicitations by about two times and a half with respect to the original design, we NEED something stronger.
Be sure, they are already even too light.