That’s true if, and only if, the matchmaking process if perfectly randomized (like a die roll). However, we know for a fact that it’s not randomized. The MM tries to create “fair” matches based upon certain criteria. Matchmaking is weighted by the MMR, which is akin to loading the die. So, by its very nature, the matchmaking process can plausibly set up matches where you are teamed with bad players more often than would be possible by random chance.
To the OP
Question about your methods. Did you play both accounts the same number of matches under roughly the same conditions (same heroes, same level of effort, similar time of day on similar days of the week etc.)?
Or did you play one account until it got to a certain rating, and then played the other until it was about the same rating?
I run two accounts too, on solo and one for mainly groups. The two accounts placed fairly close together for season 8, but ended about 400 SR apart with the solo account being lower. Didn’t play same number of matches, but tried my best using mostly the same heroes on both.
I did the same thing with CTF. My main placed at 2660, but my alt placed at 1903. Same number of matches, same heroes, same level of effort, and same time of day.
While your data makes it seem like ELO hell doesn’t exist, my data shows the opposite. Which am I, plat or silver? What those placements show is that solo queue lowers your SR by a significant amount. That means your teammates have more of an influence on your SR than your skill. Makes sense considering there are five of them and one of you.