To investigate the usefulness of stream of thought I have to rebuild one of my oldest projects, "speed-mill" speed-mill as a concept is quite simple, you build a streamlined deck around the most powerful millcards you can find.
Since many of the cards in mill are very individual and sometimes hard to measure, speed-mill needs someone statistically minded to fine tune.
Theres functional speed and irl speed to consider, and sometimes one card unexpectedly stands above all others.
To give an example, if you play hedron crab, you have to expect it to be removed at highest priority. This means it is best to play it at turn 2, to get a mill out of it, but at turn 2 you could actually cast glimpse the unthinkable. If you play minister of impediments on the other hand it will be seen as having a low threat value, so the opponent might save their removal for other stuff. They take a look at it and conclude it can only mill 6, which is true, so the annoyance factor is small. So irl you can get away with milling 6 with a minister, where the hedron crab just gets destroyed without managing to mill, or it blocks your other mill in order to achieve a result.
Going through all millcards with this mentality is hellish but deeply rewarding.
I've had designs where I've had a single dream twist in the deck because it is able to mill 6 but must not block other cards from being played, and mathematically it performed its function as a singleton, but was useless when you drew a second.
To begin with I'm just putting cards under consideration.
All of this just to test out stream of thought might be seen as insane, but I want to know exactly what can be gained by using it. (For example how fast it can thin out my own amount of lands, increasing the odds of drawing other stuff) I'm also very curious about how often it might mill 8 cards. One side effect it may have is to place dream twists in the graveyard which will bring some cardadvantage when I run out of steam. I will keep track of how many twists it can support.
In the past I've had ghoulcallers bell perform that function, which is why I know dreamtwist only works as a single, but stream of thought might change the rhythm of it all.
Sanity grinding is another card that it's been a while since I've measured it in such a contest, it thrives when there are blue mana in the deck, and in past decks I've tried it out while archive trap and chancellor of the Spires were present. At that time it never outperformed mind funeral, and I turned to artifacts like shriekhorn for the speed instead. These days we've gained startled awake, which could make sanity grinding more competitive than in the past.
At this stage I'm preparing a manabase designed to play a land each of the first 5 turns, while being able to cast spells for that mana each turn as well. Naturally this is not possible to do each turn, but I can create something that does this approximately.
In the past, I've won a shared 4th place with a similar strategy, and I've only become faster and better at constructing speed-mill.
I have the winning decklist buried somewhere in a desk but might be lured into posting it again :)