The story is part of an emerging pro-Kremlin disinformation narrative seeking to delegitimise the conduct and results of the 2020 US Presidential election.
The segment's conclusion that the 2020 Presidential election was especially dirty is not supported by expert opinions issued so far. In its interim assessment on the conduct of the election, the OSCE finds that the contest was "competitive and well managed" (p. 1), noting that "evidence-deficient claims about election fraud" (ibid.) have been refuted by "election officials at all levels" (p. 3) and "national security agencies" (p. 11).
It is telling that the report uses Russian electoral lingo to describe the allegedly widespread violations committed in the course of the election. "Dead souls" (mertvye dushi) is the established Russian tactic of padding the pro-regime vote by "enfranchising" the already deceased citizens whose names have yet to be expunged from the electoral register (p. 288n12). Independent investigation concluded, it was most recently used at the 2018 Russian presidential election and the 2020 Russian constitutional referendum, but no concrete evidence exists that this practice is systematically used in the US. See for instance here and here for detailed debunking of these claims in relation to the 2020 US Presidential poll.
In its turn, "carousels" is a crude ballot-stuffing method which involves bussing voters from one polling station to another for repeat voting. By contrast, multiple voting, whether premeditated or not, has had an infinitesimal impact on the outcome of US elections. For instance, out of more than 136 million votes cast in the 2016 US election, just over 8,000 absentee ballots were rejected because the voter had also cast a ballot in person, which accounts for roughly 0.006% of the total.
The report falsely claims that international observers were absent from the 2020 US election. In reality, an OSCE delegation has monitored the course of the presidential campaign and voting day since 29 September, and is scheduled to depart between 8 and 12 November.
Lastly, the claim that 300,000 ballots went undelivered as part of the mail-in voting process is false. Although that is roughly the number of ballots with missing delivery scans, that does not mean they were lost in transit. According to the US Postal Service, "the lack of delivery scans were a result of ballots that were picked out of the mail stream to speed up their processing, and "[t]hese ballots were delivered in advance of election deadlines."