There's plenty of M52 turbo manifolds on ebay. None of them work with a RHD BMW, but you'd definitely have more chance in a 164 thanks to the more upright motor.
On the rack and pinion conversion:
Be VERY careful and very thorough. You will read a lot of stuff about height and width of the rack relative to the original drag link, but this well short of the whole story and at least some of it is misleading.
I looked fairly hard at this with an eye to building a 142 rally car, and it got so complicated that I aborted the whole project before I started. Here's some of the problems I encountered:
A) Conventional internet wisdom says that the inner rack ends should be in line with the inner control arm pivots with the steering straight ahead, but manufacturers don't build cars like that. They all make the rack narrower so that the two pivot points line up on the outside wheel, with a degree of steering lock.
The best compromise seemed to be something that you could only work out with trial and error. I could not find anyone who spoke any clear sense on this topic - those that made authoritative statements were clearly speaking crap, while those that seemed to know what was needed seemed to have such a poor grasp of English that I suspect a lot of them were just bluffing.
B) Where a 164 inner tie rods moves backward as the steering is turned away from centre, and so too do the outer tie rod ends. The means that the inner TREs have less linear movement per degree of pitman arm movement the further you move away from centre, which is roughly cancelled out by the steering arms having the opposite behaviour.
In contrast, a rack's inner rack ends move in a linear fashion, but the outer TREs still increase in speed as they move further from centre.
At a minimum, this will create a different steering effect than expected.
C) Nobody ever talks about the effect longitudinal location of a rack's inner TREs, and their effect on the operating angle of the tie rods. This is at least as important as the much discussed vertical and lateral positions of rack ends.
This omission says to me that hardly anyone really understands the geometry of rack conversions, and also explains why supposedly thoroughly researched swaps often give poor outcomes, and why sloppy swaps can occasionally work pretty well.
D) Nobody really understands Ackerman worth a pinch of piddle. The fact that front mounted racks/boxes are supposed to be a bad compromise but the vast majority of cars with reputations for good steering have the racks at the front, says that front mount is the better option despite the "flawed" ackerman angles.
I concluded that the only way to get a good outcome was to measure the distance between the inner control arm pivots, and then try to find that measurement in a car that came with a front-mounted steering rack. Ideally a RWD one with a reputation for good steering.
Then do whatever you could to replicate the geometry of that car. Given the likely differences in track width, steering arm length, steering arm height, suspension travel, pinion location, shaft splines, mounting points, and rack-end heights, I figured that the probability of success was too low relative to the time I would need to invest.