My opinion, so open to debate (While I have had personal experience with positive displacement and centrifugal mech. S/C and single turbos, I am definitely not the last word in forced induction):
Sequential turbos, one large for top end power, one small for fast spool up and low end power - Very difficult to get to work successfully. Even manufacturers, with access to professional engineers and computer modelling have difficulty in making such systems work. eg. Liberty/Legacy
TT, S6/7 RX-7. The problem lies in switching from one turbo to the other. First, the exhaust gas flow has to be distributed prior to the change over point, robbing the smaller turbo and making it difficult to spin the larger one. It is also difficult to match the flow of both turbos at the change over point, particularly over a wide load range. Not doing so leads to noticeable dips in the power/torque curves.
Twin turbos in series - used i believe on some diesels, perhaps stationary engines or elsewhere that changes in load are not great. Intercooled between each stage, the idea is to increase efficiency over a single large compression and/or expansion (turbine) stage. I don't think that throttle reponse to changing loads is a priority. Long and complicated pipe runs would make throttle response poor and difficult to fit in a car.
Using a mechanically driven supercharger and an exhaust gas driven one, sequentially, has been done before, most notably by Lancia in their S4 rally car of the mid-late '80s. I also recall an HKS turbo bolt on turbo set up for
Toyota 4AG-ZE (factory S/C'd). Both of these used positive displacement (Rootes type), mech. driven S/C, switched out and bypassed when gas flow through the (large) turbo was great enough for it to be efficient. Change over is easier to achieve seamlessly than sequential turbo. as the air flow through the positive displacement S/C is more closely a linear function of engine (blower) speed than gas flow through the whole system and there is no distribution of exhaust gas required.
Mech. driven centrifugal S/C makes some sense on an engine that is restricted in valve area relative to piston area and therefore limited in high engine speed air flow, such as a two valve per cylinder engine; the S/C works best when the engine works worst (at higher engine speeds). Because the output of a centrifugal s/c is a non linear function of its speed, its performance is dependent on the gear ratio it is driven at. Attempting to provide boost at low engine speed by high gearing and bleeding off the resulting excess boost at higher engine speeds is extremely inefficient. (Although maybe some extra intercooling effect could be achieved by doing so after intercooling through a heat exchanger, like aircraft air conditioning).
For a street driven car, I think that simple positive displacement s/c, twin parallel or single turbo. are the most practical options. Which is better depends on what is trying to be achieved and how much money is available.
Positive displacement blowers fit very nicely into the valley of 'V' configuration engines. Airflow and therefore power delivery, is more linearly related to engine speed with a positive displacement s/c. Throttle response is very similar to a N/A engine.
While there isn't (someone may prove me wrong) enough space to fit twin turbos between the chassis rails and engine of VN-VS Commodores, I would suggest that twin turbos would have advantages in packaging and plumbing over singles on most 'V' engines. VT- may be a different situation, but I have no recent direct experience and I suspect not. I believe that there are still advantages in packaging and plumbing with twin turbos over single in a VN-VS and VT-.
If the turbos are to be sourced used, perhaps as Jap imports, there is likely to be a much wider range of suitable sizes available for a twin turbo set up over a high flow single. For this reason, twin turbos on a budget turbo RB30 may also make sense. (The issue is less clear cut because of the large (aftermarket) development history for this engine with turbo and wider range of turbos of suitable size available).
Most street driving is done at low engine loads where boost response is important = adv. for positive displacement blower. At the higher boost levels required to make really big power, blower efficiency falls away so centrifugal compressors, turbo or mech. driven start to look better.
In theory, twin turbos should have better response than a single, particularly where plumbing runs and gas volumes can be kept smaller with twins, as in a 'V' engine.
A well set up and sized system of any configuration will outperform a poorly set up system of theoretically higher efficiency.
Re: power to drive a mech. s/c. The power required to drive a turbocharger is not entirely free either as the turbine impedes the flow of exhaust gas out of the cylinder somewhat. This is less of a problem where the exhaust valve area is already restrictive, as in a two valve per cyl. eng.
