So it definitely goes off and stays off, except for return from holiday setting which you can’t stop.

During my recent call to technical I was told that, even though I have the legionella cycle turned off, it will still operate

That is, I beleive,  the biggest pile of Bullsxxt ever (or at least, not the whole story).

I have an Arotherm+ and have no legionella cycle configured - we don't need it. Our DHW runs exactly as configured, heating to 50 and no higher, exactly as set by our schedule and hysteresis. There are no legionella cycles. It's all monitored on OEM so I can see exactly what happens.

interesting you mentioned the legionella cycle.  During my recent call to technical I was told that, even though I have the legionella cycle turned off, it will still operate.  We had made a private decision (my wife is medical) because of our cycle of usage to choose a small tank and turn off legionella heat.  So the Vaillant statement confused the heck out of me

Interesting and possible I suppose.  Of course this behaviour might be firmware-version dependent.  I have never tried turning it off, but I think I will now just to check! Suggest you do the same.

I am told by another user that in absent mode the legionella cycle turns off in absent mode but then runs when absent mode is cancelled/comes to an end.  This makes perfect sense, but frustrates another user who wanted to use absent mode to turn off the heat pump for a few hours per day!

The problem with heating is that, although simple in principle, there are any number of variations in practice.  This leads to quite complex controllers, providing loads of options for householders/installers so they can (hopefully) configure it to their house.   Even my previous Worcester Bosch boiler, dating from 2005 and with largely analog electronics, had a fair few obscure options buried somewhere.  Maybe AI will one day simplify this for us, but if the AI is trained by unvetted British installers (and in fairness unvetted installers from several other countries) it will just do the wrong thing whilst making the interface even more obscure.

interesting you mentioned the legionella cycle.  During my recent call to technical I was told that, even though I have the legionella cycle turned off, it will still operate.  We had made a private decision (my wife is medical) because of our cycle of usage to choose a small tank and turn off legionella heat.  So the Vaillant statement confused the heck out of me - assuming it’s true.  Apparently the only way the cycle is limited is when in absent mode.

Mostly I find Vaillant controls good, but occasionally not.  The legionella cycle warning (hot water) seems to come on before the cycle runs for example.

On a similar vein, I “boosted” the DHW yesterday when the temp was 43.5 with the hysteresis at 5C.  The app glowed orange but the system didn’t respond with an increased flow temperature to reheat to 48C. Not understanding, I call Vaillant Technical and was told boost means, as soon as there is demand from a tap, the system will kick in.  This seems weird to me and wondered if you had the same understanding?  

No it's not and i think what yhey said is nonsense (how would ut work?)

Boost is, so far as i have been able to tell, a one off 'on'.  However I think it works only if the current temp is less than target -hysteresis.  In your case it wasn't so (in my understanding) it didn't reheat.

My confusion was when, at 22.00 and the system started to reheat, I tried to turn it off again for 2 hrs.  I set a new time period 22.20 to 23.50 with a temp at 8C and the heating didn’t turn off again.  I wasn’t “worried” expect that the controls didn’t do what I expected.

On a similar vein, I “boosted” the DHW yesterday when the temp was 43.5 with the hysteresis at 5C.  The app glowed orange but the system didn’t respond with an increased flow temperature to reheat to 48C. Not understanding, I call Vaillant Technical and was told boost means, as soon as there is demand from a tap, the system will kick in.  This seems weird to me and wondered if you had the same understanding?  

Not sure exactly what your time pattern was when it didnt go how you expected but, if you still had midnight to midday programmed at 18.5 I would have expected it to turn back on at midnight. 

Dont forget that the timer is really a daily timer, there is a sort of wizerd to program the same for several days, but the underlying data I am pretty sure is 'stored' per day.  This introduces ambiguity over the definition of midnight, which is one reason I tend to avoid time periods that start then!