1) Seats that are only D+4 or D+5 are gone in a 2010-sized wave for the GOP
2) 52-0 in California or 25-0 in New York have an absolutely 0% chance of happening because the VRA and the Democrats' diverse coalition protects maj-min inner-city districts that Democrats can't unpack. Black and Latino Democrats in state legislatures would never assent to maps that eliminated minority representation in favor of maximum partisan advantage, see: Lacy Clay, 2010.
1) As lfromnj explained, D+4/5 is more strongly Dem than you think (and also after working with the map a bit more, it seems to be possible to have all the upstate districts be D+5/6). In practice, Republicans are quite unlikely to win those sorts of seats. However, even if there is a strong enough GOP wave that they are winning D+5 seats in a 2010-style wave, why does it even matter at that point? If Republicans actually won any significant number of D+5 seats, then they already would have long since won the house by winning easier seats in other states. Those D+5 districts would simply be 1 term rentals, and control of them would never make any difference for GOP control of the House.
2) Actually, the VRA is not really a hindrance to drawing shutout/unanimous maps in either NY or CA. This is mainly because they are sufficiently strongly Democratic and have enough white Dems regardless. In addition, however, there are 2 other things to consider.
First, for black Dems in states like NY, the main threat to their re-election prospects is generally the prospect of a primary challenge. Black NY Dems are much safer in a primary if they have white Republicans in their districts rather than white NYC progressives (who may support progressive primary challengers). So actually they become safer and happier if their districts are drowning out white GOP voters. Since you mentioned Missouri, actually there was a similar thing going on in MO-05, where Cleaver wanted to have rural whites (which are more Republican) in his district rather than having more urban/suburban whites in the Kansas City metro area, because it made him less vulnerable in a primary.
Secondly, in the case of Hispanic VRA districts, it is generally good for Dems to draw more Hispanic VRA districts in states like California. The reason for this is that Hispanic population has a lower proportion of turnout to population, so you can actually draw a larger number of safely Dem Hispanic VRA districts by using a given number of Dem votes than the number of non-VRA districts with more white voters that you could draw given that same number of votes. This then leaves more Dem votes left over for other white liberal districts which can be used to drown out higher turnout white GOP areas. So the general effect of that doesn't really make gerrymandering CA harder, it actually can even make it easier in a lot of cases.
So anyway, there is no technical obstacle to drawing 25-0 or 52-0 maps for states like NY and CA, the only obstacle is the political willingness to draw the necessary lines (and, in the case of CA, to get rid of the commission). If that political willingness ever arrives though, then those maps can be pretty easily drawn. Whereas in states like TX and GA and FL and MI that Republicans control, they have to concede at least a few Dem seats even if only out pure self interest, because those states are more competitive. It is for that reason that Dems ultimately have more seats to gain than Republicans from a descent into general no-holds barred gerrymandering across the whole country.
Poll
