1) The only 2 static House districts in 2018 nationally that flipped D-to-R were in Minnesota.
This is a very good demonstration of the technical effectiveness and organizing prowess of the Minnesota DFL as a political organization. The position of the statewide house total in 2018 (D+11 points) relative to the nation did move to the right compared to 2012 or 2014, but slightly less so compared to the rest of the Upper Midwest (except IL).
2) It's an 85% white electorate that's disproportionately older and working-class.
Minnesota is definitively
very white. However, the percentage of older people (as variously defined by the census) and the median age of MN are both almost exactly around the national average. MN also has the youngest median age of all Midwestern states and the highest % of people under both age 18 and age 5 per the census.
MN is also not quite "disproportionately" working class. All 2018 and 2020 exit polls show an electorate that's somewhere between 44-48% white non-college, which means white working-class voters is certainly the largest voter bloc in the state, but less so relatively to all other Midwestern states except IL.
3) It's quite possible that Trump would have won MN in 2016 had third-party preferencing been lower.
This is likely correct. Republicans in the MN Legislature won between 49-50% of the statewide vote in 2016 across the two chambers and defeated the DFL statewide by 1.2 points in the state house total, outright receiving more votes than Hillary Clinton's total in Minnesota that year.
4) Over the past 5 years, it's been readily proven that high turnout is no longer inherently 100% good for Democrats: a 100m+ electorate in 2010 or 2014 likely would have seen Republicans routed by >10 points. Many would have expected the same in a hypothetical future 2020 presidential with ~160m voters.
High turnout in MN indeed doesn't seem to help one party or the other. Over the past decade, the Democratic base in Minnesota has primarily been secular white voters with college degrees, not racial minorities, although this is beginning to change. White voters with college degrees in MN backed Biden by nearly 30 points in 2020 but only voted Democrat down-ballot by around 15-20 points, and their "elasticity" when voting on the same ballot certainly seem to validate your point.
5) While the strength of Democrats in the state has been held together almost exclusively by the remarkable local influencing power of the DFL, polarization, generational turnover and a lack of interest in day-to-day civic participation means people are less likely to cross party lines for any contest whatsoever.
The post-Trump era view of the various campaign committees about MN's partisan politics has been that the DFL has indeed traditionally overperformed relatively to the state's expected demographic baseline, but that the said baseline itself still retains a small Democratic tilt even when state party effectiveness is removed as a factor. The DFL's local "ancestral" strength in far-flung exurban WWC areas began to crumble in 2016, continued its demise even in 2018, and has largely evaporated by 2020 except for a few residual state legislative seats in the "Iron Range" that the DFL is guaranteed to lose this upcoming November. It's also worth remembering that rural counties in MN actually voted slightly more Republican in 2020 than rural counties in neighboring WI and about as Republican as rural counties in IA, both states that were more Republican statewide than MN.
The fact is that Democrats could actually get a majority in both houses of the MN legislature solely by holding all of the Twin Cities' second and third ring suburban districts without winning a single district anywhere else in the state (though only with a bare ~1-2 seat majority). Furthermore, if things break down to purely national trends of urban/suburban vs rural/exurban, the Minneapolis/St. Paul core and inner suburbs also has enough population to keep the state at a minimum competitive in federal statewide elections for years to come.
