Personally, I would wonder a bit about OCR. I would avoid obvious words or phrases. For example WWG1WGA could be easily flagged by OCR I would think. Q hasn't said this, I don't think, but it stands to reason I think. Then again the cabal uses a lot of the same words in their memes (which are probably greenlit by the same algorithm that flags our images) so… Not sure. Trying to re-phrase the words to have multiple meanings (the way Q does to bypass filters) might be a good idea anyhow.
Actually I notice that a lot of the Fake Anons on twitter just use the same memes over and over. These memes are probably pre-screened so that they get through the algorithm. They're probably put on a whitelist, so to speak.