However, based on the data you obtained, I still think that the main influence is given by the STEEM price. The burning of the SBD or STEEM, on the other hand, has only a minor influence.

Yep. I completely agree. In total, I think we have burned somewhere around 6K STEEM, ⁶/₁₀ of a percentage of the four week difference in virtual supply, so I guess that >99% of the reported 4-week change comes from STEEM's price change.

According to information from coinmarketcap, "circulating supply" means that only free tokens are included.

My theory also relies on this point. I need to look at the numbers over the course of a couple weeks to see if it holds up. Unfortunately, I'm not aware of a way to get the historical numbers that I would need in order to check it using past data.