TVS superchargers are only mapped to 20psi as that's the maximum design pressure they were engineered for. Mapping a supercharger requires the unit to stabilize at each point before the data point is taken which can take up to a couple minutes so map data isn't available above these points.
Now that being said I don't think you're numbers are far off, but the calculation for RWHP doesn't work like that, it's just not that simple - although the turbo manufacturers would like you to think that... below is a note that a wrote a while back relating to the LS9 SC, same principals still apply to the 2650.
It is important to note that this input power requirement should not be confused with the parasitic power requirement of the supercharger. This always seems to be topics of discussion around different type of boosting devices and is normally completely misunderstood.
First of all the isentropic (or adiabatic, or thermal) efficiency drives the power requirement of the compressor excluding mechanical efficiency. In other words, any type of compressor that has an equivalent efficiency will have the same input power requirement (including exhaust driven devices). No device is free from mechanical losses and most are quite similar so this aspect can normally be overlooked for comparison purposes.
Parasitic power is simply defined as “power that is being consumed without any useful return”. In other words, pretty much anything that exists has a potential for being subject to parasitic power. If a supercharger is working to provide increased manifold pressure for the engine, that is not parasitic power, it’s work is providing a useful result – increased manifold pressure.
To understand parasitic power of a particular boosting device you need to look at the entire boosting system including it's interaction with the engine.
At vehicle cruise conditions for any boosted engine, increased manifold pressure is normally not required so any consumption of power from the boosting device is considered parasitic. In the particular case of the LS9, the parasitic power consumption of the supercharger under cruise conditions @ 70MPH would be approximately 0.86 HP. For exhaust driven devices the parasitic power is calculated from negative pressure differential across the engine (increased exhaust backpressure - which normally results in about the same value).
At WOT the parasitic power calculation is a summation of the frictional power requirement of the supercharger (which is a function of SC speed), system pressure losses (normally driven by packaging), and the isentropic efficiency of the compressor - this calculation applies to all types of boosting devices, including exhaust driven devices. As a side note, the R2300 supercharger at peak power on the LS9 has a frictional power requirement of 1.2 HP which is defines as the losses through the bearings, seals, gears, and oil.
For a mechanically driven supercharged application, many people seem to quickly point out the amount of power required to drive the device. This is primarily because it is easy to measure and is normally published information - but this is not the amount of power that needs to be paid for by the engine. Mechanical supercharging places the engine at WOT under what is referred to as a "positive pressure differential" - the intake manifold pressure is always higher than the exhaust pressure. This is an important factor as the engine, in simplest terms, can be considered a positive displacement compressor. When a positive displacement compressor [engine] is subject to a higher pressure at the inlet than the outlet, it becomes a motor that is capable of providing power. This is the condition that the boosted engine operates under with a mechanical drive supercharger, and when properly optimized, the positive pressure differential will provide a certain amount power output from the engine - this can then subtracted from the required input power.
As for peak power numbers, the highest I've see for a TVS 2650 was right around 1490 RWHP on a Patterson Elite built 427 LS, that car went 6.7s at well over 200MPH.
Displacement and RPM are your friend, Not sure how high they spun that 427, but since it was based off a COPO and they are rumored to spin over 10K rpm, I can only guess it was spinning pretty good also. COPOs are believed to be just under 1400 RWHP running 7.8s at nearly 180 MPH with a weight of 3500.
Hope that helps.