Players might not understand what the spinners look like in the editor, but I think they will realize that the banana density is higher than normal in these difficulties.
In my difficulty, for example, you can see the difference in density at 01:34:551 (1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1). Here, I'm using a spinner stream (lots of bananas) to represent a series of glitchy sounds, followed by a normal spinner (fewer bananas) to represent a long note. You can also hear the difference between both spinner patterns: every time you catch a banana in a normal spinner, the pitch increases, whereas the pitch is constantly reset in a spinner stream, which creates a glitchy effect that fits the streams in this song. Since I'm deliberately using both spinner patterns in my diff to represent different sounds, I think it would be a shame to remove one kind of spinner pattern in favour of the other. I did not have score and object count in mind when I mapped these spinners, and I don't think these values matter, considering they don't affect pp or star rating.
For all of these reasons, I'd rather keep these spinner streams
Concerning the kiais, I added curves at 01:06:293, 02:13:261, and 02:30:293 to remove some zigzag patterns. As for the calmer parts, I added more standstills around 00:14:035, 01:15:970, 01:53:132, and 02:17:906 to reduce strain.
01:54:621 (1,2,3,4,5,6) - Added more standstills. I only kept one antiflowy 1/4 hyper at 01:55:728 (4,5) to emphasize the very loud sound on the blue tick.
02:08:312 (1,2,3,4) - This is the song's climax, I think the antiflowy 1/4 pattern is justified.
02:33:279 (1,2,3,4,5,6) - Added more standstills here as well, but I kept some antiflow.