These are the FiveThirtyEight model forecasts for the six races that Democratic groups ran ads about the MAGA candidate in the primary. Gibbs is the only one with a real chance. Looks like the strategy may have paid off.
Gibbs defeated Meijer in the primary. This was Justin Amash's old seat. MI-03 was Republican-leaning but the district got changed in redistricting and became more Democratic-leaning. Therefore Democrats decided to spend money on this primary in the hopes of picking up the seat.
This was a controversial strategy, but I think it said more about the state of the GOP than anything. The Dem ads were often just highlighting the extreme stances of the candidates such as not believing Biden was legitimately elected.
Of course the Dems are highlighting their insanity knowing the Republican primary base loves that insanity, but again primaries are decided by primary voters. This strategy only has any purchase because the base is insane. In other times we wouldn't expect an opposing party to save a party from itself.
I think there are some pretty significant differences doing the strategy with presidential races compared to others. The electoral college for one. People are more likely to vote outside their party for down ballot races too. Even a clunker of a presidential candidate has a hard floor of support just by virtue of their party identification.