I find that in deed you can pair certain reverb types to certain sound sources, but you should also not do so.
My taste is that thick sound sources should have thin reverbs and thin sound sources should have thick reverbs. A spring reverb is a thin reverb and that suites well on rather fat sound sources such as organs and pads, to give them dimensionality without providing a lot of mud to the mix. Thin reverbs I like to be present on all volumes, thick reverbs I want to exceed the perception threshold when you increase the playback volume to rather loud volumes, to get additional impact when you turn up the playback volume. Then, depending on the mix of sound sources in the production I will look at what is a good mix off reverbs on these, but usually when it comes to the thick reverb I tend to assign that mostly just to the snare, especially if it is a dense production. Some will be added after the delay, some on the actual hits, in a dense production I might just have a fat reverb on some hits on the delay of the snare. Usually it is the spring reverb that dominates the mix, but they are so lightly added that you can almost not hear them and many of them are disabled in the verses, I don't like it when there is a ping of the spring reverb present. When it comes to more fat reverbs, it is mostly strings, violin and those kinds of sound sources that I add reverb to. Sometimes it can happen that I choose to thin out the acoustic guitar a little and then apply a fat reverb onto that at a very dry setting, just to help give the mix some added resonance in the stereo field, but when I do so I typically use a Bricasti. I am not a great fan of a lot of reverb on lead vocals, but it does happen that I use a fat reverb on the vocals, I ensure it does not slap into the mix at the same time as the snare. But when I add reverb to the vocals, I might have side chained an offending frequency range so that the there is room for a quick perception of the reverb. I am extremely careful about that since too much fat reverb on vocals can easily become pretty muddy. Sometimes I want a bit spring in there as well, it depends a little what suites the particular singer and song. The reverb time I usually want very short if I am using software. With hardware I want a bit longer reverb but sometimes also a short reverb works. It happens that I put a Bricasti on the whole drum kit, because the Bricasti has this character of bringing out a certain sense of "live" vibe and sometimes that suites the mix very well, especially when there is a lot of room for the drums in the mix, I might even decide to pan things a bit wider than I normally do when I add the Bricasti on the whole drum kit. Hardware Lexicons tend to have a size increasing impact, so I use Lexicons for adding size, typically on vocals/background vocals perceived when you turn up the playback volume. For electric guitar I tend to like delay more than reverb, but it totally depends on the production. In many cases the guitars play a rhythm enhancing role to the mix and then I don't add any reverb to the guitars, instead I ensure the reverb leaks into the perception from some other sound source. Typically I use a delay on it instead timed to add an increased rhythm perception. Although I do all of this, I also do the opposite to bring excitement into the mix, but I do so cleverly. I don't stick to either one, to bring enough dynamics into the mix I want both application types in there. Both the delay and the reverb I kind of map out early on, it is a matrix on the mix, it does not sit there eating up the signal, revebs and delays are some of the tools I use to bring dynamics to the mix. Instead of increasing the wetness I might just unmute various types of the same kind of reverb (but a different fx) using automation, it's just that I do it with hardware, that's the difference, it stays clean but adds dimensionality/width/depth. With software that same depth/width results in a less present sounding mix, which depending on various things can damage the mix a lot. This is one of the reasons why many ITB engineers struggle with the "loudness" of their mixes, they have simply too much total reverb in the mix and too long total compressor release time, often with a gain staging issue present as well that makes it so that it does not matter what you do on the mix bus you will always lack tons of presence compared to a commercial counterpart. If you are using an all ITB platform, it can pay off to limit the mix to the target loudness before adding reverb and delay, so that you are at least aware of when the presence is lost, because that is something that is unique about your particular ITB setup.
It is good if you can bring in a lot of natural reverb during recording, so that you don't become focused on that during the early phase of the mixing process. I tend to squeeze out the room reverb the best I can from the recording first, then I add reverb and delay in mixing and mastering, mostly just for the dynamics and how the mix grows when you boost the playback volume. Much of my time during mixing and mastering goes into the analyzis of what happens when I turn the playback volume knob. If you have a great density charge in the final master, it is awesome to turn up the volume and the reverbs and delays play a critical role in that since they can be resonance enhancing. When you add the right hardware reverb and delay into the mix in a good way it can really make it alive.