Cope elimination is a pericyclic reaction that always prefers syn-elimination. Due to hydroxylamine being a poor leaving group, the reactions often tend to prefer hofmann products.
However the reason for preference towards hofmann products is the stability of carbanion formed due to less $\ce{\alpha-H}$ and consequently less destabilization by +H effect. (since leaving group is poor, the acidic-H departs before the leaving group departs inducing negative charge at the $\ce{\beta-C}$)
The carbanion's stability is enhanced by the resonance effect in your first and last examples, and we know that resonance effect is often more stabilizing than hyperconjugation hence the preference of zaitsev product. Resonance effect almost always gets preference even if the reaction often produces hoffmann product. (since the leaving group can always be more bulkier there are edge cases)
Regarding your 2nd and 3rd examples they are probably giving zaitsev out of the temperature range for hofmann product in cope elimination. Where resonance effect is not in play, the determination between hofmann and zaitsev becomes a balance between the thermodynamic stability of the final product and the kinetic stability of the carbanion transition state.
All of the above examples at very high temperature will provide zaitsev product as thermodynamically favored product is formed. However in a certain temperature range these reactions give the hofmann product where the kinetic stability is dominating. Cope elimination produces hofmann product near $ \ce{T \approx 100^o C}$ pyrolysis of xanthates produces hofmann near $\ce{200^o C}$ and esters give pyrolytic syn elimination near $\ce{500^o C}$.*
The conclusion is that unless temperature is given, it is not exactly clear that you need to produce hofmann which is also stated in wikipedia
There are many factors that affect the product composition of Ei reactions, but typically they follow Hofmann’s rule and lose a β-hydrogen from the least substituted position, giving the alkene that is less substituted (the opposite of Zaitsev's rule).1 Some factors affecting product composition include steric effects, conjugation, and stability of the forming alkene.
The example of 5 member T.S. in wikipedia also supports zaitsev products. Basically the selectivity is not very good, and for correct answer we can only refer to data given to us. This is why temperature is important to mention.
*Source: Peter Sykes 6th edition section 9.9 (Pyrolytic syn elimination)