So I tried to remake one of my RCT3 coasters in RCT2 which was the Heartstopper, and if you look on that page, you'll see that both the excitement and intensity are very high. However, when I tried to remake it in RCT2, the excitement came out as low as the intensity came out as ultra-extreme, which is usually way too intense for guests. So does that mean that the testing results logic is different between RCT3 and RCT 1&2, or is there something I'm missing here?

RCT1&2 used a combination of excitement and intensity to make riders want to go on a ride. RCT3 changed it to make nausea and intensity factors that drive guests away while excitement draws them in. So yes the system was reworked between the 2D and 3D titles.

RCT2 uses the intensity rating to determine if a coaster is safe by realistic standards. If a coaster's intensity is ultra-extreme, that usually means the lateral Gs are too high or the train barrels through an inversion too rapidly. Since excitement and intensity are inversely proportional after an intensity of 10 is reached, almost every roller coaster with an ultra-extreme intensity rating will have a low excitement rating. 

RCT3 nerfed the rating system to allow roller coasters with unrealistic G-forces to still be rideable, which defeats most of the purpose of the intensity rating. So you can build dangerous coasters with off-the-charts lateral Gs in RCT3 and guests often won't complain at all, but try and build the same coasters in RCT2 and nobody will go near it.

^ Thank you for clearing up what I tried to say Terry

Yes I think that from this point of view RCT1 and 2 are more realistic than 3.

Thanks guys