I feel if I have a dyno stand in my shop and include a graph of engine power it gives an added level of credibility to a build I do. Many guys say they want a 750HP electronic engine. I try to explain a 400HP 12 valve is a pretty wicked amount of power. More than most people would want and way more than you need. I take them for a drive in a 300-350HP 12 valve powered vehicle, they shit their pants and forget all about those big stupid numbers and I get paid.
That's why I brought up the bullshit numbers many aftermarket dyno builders produce. As you say, "take them for a ride and ..." but that's another wasted layer of explanation
The other purpose is I'm involved in an opensource CanBus development project. We're not far enough along yet to know what we need exactly, but there's a good chance solving a few mysteries is going to involve forcing conditions on a vehicle drivetrain while sniffing things out.
This is why I think you'll eventually want better control ... continued below
I don't follow the accuracy part. Dyno is just a load. You can't fake that.
Yes you can, and they do. Not that it matters for you too much, you'll be using the same dyno for everything you do, so its comparative and you probably don't care too much if Customer B badmouths you because your "dyno" numbers are lower than Shop C in Omaha that produces 2000 hp duramaxes ... just something to be aware of.
I guess the part where things are fuzzy for me is how you would best utilize a manual dyno- You have one hand on the throttle and one hand on the valve control- To get real numbers. My friend that ran old Stuska's says they just did it incrementally every 500 RPM on gas engines to plot points and connect the dots. I see that being iffy to do with a turbodiesel, but I have no real world experience.
What people did in the old days with manuals was what you describe - make a change to the motor, stick it on the dyno, run it at various rpm and record the results. Compare charts. Even then, this wasn't the complete answer. During weeks of testing on Ax'es Big Red, Mert and Parra would take a pair of XR's out on the (public) road and do roll-on tests. Manual dynos don't do acceleration testing. Marc, where I sent you, ran a manual dyno for several years, modified it in an attempt to get more data, and finally built and sells his own eddy current dyno now. You can get so much more useful info out of a testing device with more advanced controls.
Sure, you can gopower as a simple test stand and that's probably a good start, but the more you know, the better you can do, yes ? If you are going to start putting money into it, then having an idea of where you ultimately want to be could mean less time and money wasted up blind alleys.