The mechanical part of dyno's is pretty much all the same. Either weighted rollers or a stand that attaches to your cars hubs. Then your car spins them and the acceleration is converted to hp. Dynojets spin I think like a 1000lb drum and there's not much you can manipulate so they're pretty reliable for hp gains and such. Only downside is that if you tune on a dyno, you need to compensate for the street or track since the load is higher. So a safe AFR on a dynojet may be lean in real life. Mustangs spin a drum as well, but you can change the load to compensate. It's great to be able to adjust for things, but you can also make them lie. Input a lighter load or different car and all of a sudden, you have a 400hp stock Civic. I wouldn't put much faith in a Mustang for say a manufacturer's claims about performance gains for their intake. You can skew the numbers so much.
Then you have Dynapacks that attach directly to your hubs. You remove the wheel and then bolt stand to your hub. Supposedly takes out the variance in different wheels. In reality, a lot of people just like them because they read high. Great for internet bragging. Bottom line about all dynos is to stick with one type. They all work well, but if you get tuned on one type, then it would be best to stay with it otherwise you really won't know your gains and what's working and what isn't.
These days, it's easier to dyno an automatic. Just pick either 3rd or 4th depending on the car and you can usually use the manual shift mode to keep it there. Back when I had to dyno my automatic, there wasn't a way to lock it in gear so we could only get readings starting around 3k rpm. Otherwise, it would kick down a gear. As long as TC is locked it's not any different than a manual. We also had to make sure the car didn't upshift so we had to back off the throttle right before. Really don't want a full throttle shift on a dyno.