To get back on track, the reason transaxles are so expensive is
because they "CAN" be.
This has always puzzle me too. Companies have been making gears and the other associated parts that are in a transaxle for years so there is no need for any sort of special equipment to manufacture these things and putting the transmission and differential in the same case shouldn't have anything to do with making them more difficult to manufacture. The only thing different here is the volume of them that is manufactured. A few dozen verses many hundreds or even thousands seems to be the only reason for the high prices.
No kidding.
People don't bat an eyelash at paying $2100 for a Tremec TKO500 transmission and then another $1850 for a built Ford 9" rear. Transmission and differential, sort of a requirement for a car.
The transaxle is both transmission and differential all in one and is produced in quantities of probably 1/10000th of that the TKO500 and Ford 9" rears are. Along those lines I think $5000 is a relative bargain for a transaxle. And, rebuilt G50s with LSD are available for about that, $5400.
Doesn't seem like they are really all that high to me.
So, suddenly profit is a dirty word in the USA and those engineering companies making highly specialised precision products for an extremely small and volatile market are just plain "greedy?"
If you can't (won't) afford whats available in the meantime, perhaps your cash might be better invested in a model train set.
I have to fall into the baffled category. If I'm not mistaken the original T44s were based around the top loader gears. That would cut costs signicantly, I would think. A case is needed and a shaft or 2 to "unify" the components in the case...but DAMN. It would seem that the longer it takes to make your "goal" to break even, the more they will cost.
Why is it, as mentioned with the Tranzilla analogy, that internals can't be an existing, proven set. That would cut costs enormously and have a proven record. I guess I'm a tightwad, I think anything over 10K to 12K is at the top end of reasonable. The advantage I see over using existing gearsets is the readily availability of replacement parts.
I recall when working as Benelli wrench in the early 70s that they built a 4 and 6 cylinder bike that was based on the use of Honda 500 4 cylinder parts. That made parts access, which was always an issue with small manufacturers in the US, a non issue. So I have a little trouble understanding why a similar approach can't be done with a transaxle.