[TurboKraft]

The bummer about the Garrett offerings is that they don't have many non-wastegated turbine housings in D5 or cast stainless.

True, stainless direct from Garrett would be nice, like the sexy Tial v-band turbine housings. Or better, the Inconel housings of the ultra-cool TR30R (Audi R8 Le Mans turbos). Garrett uses high-nickel content "Ni-Resist" alloy for their GT-series ball bearing turbos. Not a metalurgist, but it seems incredibly tough and withstands the EGTs of these air-cooled engines (and EGTs of rotary engines, also notoriously hard on exhaust parts).

So when somebody makes a "hybrid" turbo by screwing a compressor to a turbine, to me it is like a house painter retouching the Sistine Chapel ceiling.

Another factor is the rotational inertia of the rotor (Compressor + Turbine + Shaft + thrust bearing). The higher this is, the harder it is for the turbo to spool up. I think you will find that newer turbo's have much lower (30% lower) than turbo's from the '80's. That is what 25 years of computational fluid dynamics buys you.

Couldn't have said it better.
Long story short, the engineers at Garrett (and BorgWarner, etc.) have the tools to refine their products very well before putting them into production. They don't match one compressor wheel to a different turbine wheel by chance or accident. They look at the airflow characteristics and inertia of the components and put together the best combinations. Rarely will you find a manufacturer's turbo with a compressor wheel that radically out-flows the turbine wheel, as you often find in the aftermarket. (And these are usually targeted for small displacement, high boost, high HP engines that also run massive wastegates to make up for the lack of turbine flow.)
That's not a 930 engine's m.o.

One result of this mis-matching is higher exhaust backpressure at the turbine inlet (header) flange. 930s are bad enough, and the long, undersized tube feeding the wastegate on most aftermarket headers doesn't help matters, either.
Anyone else have exhaust backpressure data they would like to share?
We just dyno tuned an ex-Ruf 3.4L that had been in storage for many, many years. Another shop converted it to EFI and "600hp" -- not with a K29-7006#13, buddy! At 17.2psi boost in the manifold, the exhaust backpressure was just a hair under 30psi. About a 1.75:1 pressure ratio (backpressure:boost) despite the larger #13 turbine housing. Car is choking on the backpressure, spooling up slowly, keeping the Group B cams from doing their work.
One of the most responsive EFI 1994 Turbo 3.6s we have built had 2 different large ball bearing turbos. The first had a 1.4:1 backpressure ratio and drove great. The next turbo we used had a lower 1.2:1 ratio from a significantly larger turbine wheel -- and it spooled up even sooner than the smaller turbo!

The reason no one looks at a turbine map is that it's really trivial for the end user to do so. The end [user?]sizes the unit using a/r ratio, trim and max speed of the turbine at a desired compressor output (which conveniently appears on the compressor map). Looking at the map that you posted, what could we understand that those three things don't already tell us or how that map overlays a compressor map?
Only guys like you use turbine maps to impress people at parties and on forums.

Have to disagree. We use the turbine maps a great deal in making an initial turbocharger selection on a new engine combination. If you know the cfm flow of the engine, you better have the exhaust flow to match otherwise you get choked with backpressure into the turbo. Sure, we might make a change after initial dynos, like a different A/R turbine housing to tailor the power band for how the client will use the car. But this turbine flow data is very useful in selecting the closest match the first time around.

Now... Garrett GT35R guys with CIS- How much boost can you run before the CIS runs out of steam?

So far, on a <8.0:1 3.4L, we're done at 0.75bar because AFR is approx. 12.5:1. On this west coast pump junk they call 91-octane, we'd rather be conservative and keep it just richer than 12:1.


What's the most power anyone has made from a K27?

The most I've ever seen was 455hp from a 930, and that was on MODE's (Holcombe's) engine dyno. 3.5L 930 with all usual bolt-ons and cams, and a K27-7006. It made the same power at 1.1bar as at 0.9bar -- 455hp -- only at lower rpms as boost was raised. End of the turbo, turn the boost back down, no point raising boost when the turbo is out of air.