This does seem to be a bit of a nightmare to be honest .. there's many values i would like to see .. i'll just humbly suggest an Indy has a look at this point as it may be the easiest solution .
Either way i'll run through what i think atm .. this might get complicated .. sorry .
You said every few drives .. do you know how many ? .. basically a fault with secondary air should logically be thrown up every cycle .. cold to hot running .. say 10 miles .
Lambda sensor or emissions related faults often take several cycles to flag a fault .. i tend to say 4-5 for a fault to be reported ... if several cycles then it leans towards the lambda .
My confusion though .. if it was only codes for bank 2 then its a lambda fault or something related to that .. air leak , partially blocked cat , cracked exhaust .
Secondary air on B1 is a seperate fault .
What i'm thinking from this .. is B1 lambda almost at the point of triggering a code but not quite .. ie we have a running issue but atm only B2 lambda is throwing a fault code .
Values needed to have a clue ..
B1 and B2 .. rkat , Frau and Frou .. your tester may indicate them as adaptations idle , mid range and max rpm .. its the fuel adaptations the car has to make to maintain lambda on each cyl ... ( lambda its the correct air / fuel mix )
maf readings at idle .. kg/hr hopefully .
Watch the live readings from both bank lambda sensors ( engine hot and at idle ).. you should see a switch between 0.2 and 0.8 volts .. up down , up down .. do they switch at a similar rate .. if you have data logging then graph both and photo it please .
I need to know if B2 lambda is switching or staying lean ... if it stays at 0.45 then its dead or the dme is not getting a signal from it .
Fault code says dynamically inert .. that to me means its not switching .
secondary air pump is working .. it blows air through holes in the head into the exhaust manifold .. it's possible the ports in the head are blocked .. i can't see it and there's not much you can do about it but its something to bear in mind .
There is also a valve i think . not sure . you need to investigate that side .
So .. in essence , i'm thinking atm ..
Is the wiring from the lambda sensor to the dme all ok ..
Is the cat that side partially blocked ..
Is it actually a running fault but the other side is just about within limits.
Is there a crack in the exhaust , perhaps you can't see it .
Tank vent , carbon canister .. these affect both banks but it is listed in the code description .
Without looking at a lot of things , driving the car etc then i am at best guessing ..
What i will add .. if your keen and want to get involved then i might have some basic training info on these i can send you .. pm me an email addy if so and i'll see what i can dig out .
Many years searching the internet has found me the odd bit of info that's useful .. sometimes lol .
also just to say .. read this many times as there's a lot here to get your head around as after all .. your the one fixing the car and perhaps learning as you go .. welcome to my world