adjustable low v trigger 12v relay

Use a programmable microcontroller with a built in A/D converter.  You can use a 30K ohm resistor between the TPS signal and the chip which should eliminate any chance of modifying the signal.

You won't be able to diode isolate the signal and have it read down to 0.2vdc because of the voltage drop of the diode.  The diode would also modify the voltage going in to the processor.  If you want isolation (which you don't need as long as you use the resistor) you would want to use an op amp set up as a voltage follower using high impedence resistors.

You could then, using the same processor, wire up to pots and independently set on and off thresholds that would be adjustable on the fly.  You could also use an ext. int. to monitor RPM or any other signal in the car and control the output based on other signals.

You could also go about doing this with op amps alone set up to compare voltages.  It wouldn't be near as pretty or flexible, but it would work.  You would feed your TPS voltage, through a 30K resistor in to one input of a comparator.  You would then wire a pot to the other input.  The pot would dictate the switching threshold.  You would need to set up some sort of latch for the hysterisus - you could possibly use the "ON" threshold to latch a relay and then the "OFF" threshold to unlatch it.

Op amps are really easy to work with, as long as you know what you want to do with them.  A voltage follower circuit is about the most basic - it takes the voltage in and outputs it - the "front end" typically is high impedence and the output can be transistorized to make it a great tool for isolation.

An op amp in comparator mode isn't much more complicated.  You can find schematics for both with google searches.

For the "latching" I would think you can have the low threshold provide a (-) output to the relay.  Then, have the high threshold provide the + for the coil.  Have the relay "latch" itself as well as provide a +12v output.  This way, once the TPS is above the low threshold the relay will have a (-) signal but not energize because there is no +.  Then, when you exceed your high threshold you will have a + voltage for the coil, causing the relay to energize.  Since the relay is self latching it will keep the relay latched even if you drop below the high voltage threshold.  However, if you drop below the low threshold the coil will lose it's - side and thus the relay will drop out and break the latch.  It seems pretty simple with not too many parts.  You can get one chip with dual op amps which will reduce part count, then a few pots, and a few transistors, a resistor or two, and a relay and you're in business.

I'm pretty sure there are companies out there that make adjustable throttle switching - especially for nitrous applications.  I've never bought one so I don't know how adjustable they are, but I'm guessing they are made.

What kind of exhaust valve is it?  Does it have a built in stop? 

Your diagram seems like it is a bit excessive, and I don't have a full understanding as to why you are using the relays you are using.  What's the point of using the timer relays with the relay latch?  One seems like it would cancel the other out.  Also, I think the momentary out relay will mess things up - I would eliminate it and just use the (-) when on output to do what you need to do.  Since you haven't gave too much information about the application and keep throwing in new requirements it's hard to say what will or won't work.

My assumption when you said exhaust valve was something like a solenoid valve - as long as you hold votage on it it is open and when you remove voltage it closes.  What type of valve are you working with?