You do realize that now that there are 60+ posts on Q's board, you will be able to arbitrarily do this with EVERY SINGLE TWEET AND POST.
It's basically just tea leaves unless you can uncover a specific methodology that doesn't lend itself to finding dozens of false hits.
I haven't really kept this 100% updated, but it gives a good idea of all the various ways [markers] can work.
Try to take these examples all in, then try to apply similar methodology moving forward.
New methods appear all the time, for instance the post number markers (lower right corner).
You need to be sharp enough to tell the difference between a genuine new confirmation, and just a one-off biased "confirmation." Typically this means being able to show how not only have you found a correlation, but if you apply the same methodology it will only produce positive correlations. If you apply your methodology and it produces a lot of unconfirmed or seemingly unrelated matches, then it's likely invalid.
Even if you KNOW a post and tweet is related, you still have to go through this process to find the actual method of correlation.