Remember that the house incumbents who lost in 2020 were mostly moderate/conservative members who were far apart ideologically from Sanders. This would negate the notion of Sanders having a greater negative effect downballot.
If anything, doesn't this lend further credence to the fact that down-ballot Dems would have done worse, or at least as poorly, with Sanders at the top of the ticket?
The Dem members who lost (or nearly lost - in the case of members like Spanberger) were some of the most centrist members in the caucus, yet (in large part due to the GOP or at least split lean of their district, further exacerbated by Dems' lack of in-person canvassing or events) they still couldn't overcome highly effective GOP messaging (or fear mongering) as to what a Dem trifecta might entail (socialism, open borders, defunding the police, etc.) With an actual self-identified democratic socialist at the top of the ticket, fending off these attacks would have been even more difficult - and you'd probably see a few more of the tight races (Ossoff v. Perdue, NJ-07, IA-03, IL-14, VA-07, maybe PA-17 and MI-Sen) land in the GOP column. Maybe (probably?) not enough to flip the House entirely, but likely enough to keep the Senate in GOP hands.
It's also unclear to me what the impact would have been in races like FL-26, FL-27, TX-15, etc. On the one hand, Sanders outperformed Biden with Hispanic voters (at least in the early contests) in the primary, so there's a case to be made that he would have also done so in the general (which could have had down-ballot impacts), but on the other hand the socialism attacks that hurt Dems with certain subsegments of Hispanic voters (namely Cuban-American/Venezuelan-American voters in FL and Tejano voters in South Texas) would have probably had a more pronounced impact with Sanders at the top of the ticket.