Nature would always prefer less lethality and greater transmissibility. Sometimes more lethality is the price for greater transmissibility, and other times less transmissibility is the price of less lethality. It's the sum that matters.

So if the ability of a virus to transmit is environmentally compromised (say by lockdown) then a mutation which is more lethal yet confers greater transmissibility could be a more successful reproductive strategy than simply lowering lethality.

That's what I say. There are two theoretically orthogonal properties, lethality and transmissibility, that mutations can act on. In the real world, the properties might be coupled in that a mutation can improve or worsen one or the other or both. Where the mutation changes both properties, it is the practical sum total of change to both properties that determines whether the mutation goes on. Basically nature is an annealing machine.

It certainly doesnt mean *less* lethality which is what you claimed repeatedly. It could be positive, negative or neutral.

Of course the prevailing pressure is lower lethality, since dead hosts are in the equivalent of a permanent lockdown, they certainly can't go about infecting others.

Fon you're missing the point of Splatt's (correct) reasoning here. The driving pressure on evolution is reproductive success within the given environment. Any mutation which confers an advantage here will tend to be preferentially selected. So if the ability of a virus to transmit is environmentally compromised (say by lockdown) then a mutation which is more lethal yet confers greater transmissibility could be a more successful reproductive strategy than simply lowering lethality. In fact simply lowering lethality, where transmissibility has been environmentally compromised, may offer little to no selective advantage - especially in the case of a virus, like covid, which does not have a particularly high mortality rate for its hosts.

Right. So this low level lethality victimising the old is nearly irrelevant to reproductive success and may still be irrelevant at much higher levels of lethality. Therefore discriminating against the present low lethality Covid which victimizes less valuable souls like me has unknown consequences which may include leaving the field free to more transmissable critters which may also be more lethal and/or dangerous to the young.

If this is true:

1. 'normal' Covid confers immunity to the S.African variant, and
2. the AZ vaccine does not, and
3. the AZ vaccine confers immunity to 'normal' Covid, as it may,

then isn't the AZ vaccine artificially selecting for the S. African variant?

Not quite. The E484K mutant has been responsible for documented reinfections in Brazil ( https://virological.org/t/spike-e484k-mutation-in-the-first-sars-cov-2-reinfection-case-confirmed-in-brazil-2020/584 ) and the Novavax trial in SA had reinfections *in the placebo group* ( https://www.bmj.com/content/372/bmj.n359 )

Its a mutation in an area commonly used for antibody binding and also an area the vaccines generally try to mimic. So its evolved as a more general immune evasion which as luck would have it also evades vaccines that target the same place.
This is one issue with new vaccines targeting a specific area of a virus rather than older style inactivated which expose the immune system to the whole thing.

Ultimately all the vaccines showed a 6 fold or so reduction in neutralisation but because they started from an extremely high > 90% level still showed good efficacy against this.
Oxford, starting from a far lower 64% or so because useless taking into account that decrease.

So its a specific problem for ONE vaccine - the others can cope for now well enough.
The problem is the UK is basing almost its entire response on the Oxford/AZ.

It certainly doesnt mean *less* lethality which is what you claimed repeatedly. It could be positive, negative or neutral.

I should not have to teach you this: Of course the prevailing pressure is lower lethality, since dead hosts are in the equivalent of a permanent lockdown, they certainly can't go about infecting others

No. You miss the key part.
With SARs2, the infectious period is relatively small, typically a week on average at most.
People on average take 3-4 weeks to die. They're not infectious for the last 3-4 weeks.
So selection wise its irrelevant. Once a host is no longer infectious theres no selection pressure there at all. You can discard it, throw it in the bin. It has no effect.

A virus has a benefit increasing the time an individual is infectious but after that it really doesnt matter what happens.
If changes to increase infection end up doing more long term harm after that period it doesnt matter - its still a benefit.

Of course every disease in history does spend its time attempting to get more and more transmissible.

It doesnt "attempt" to do anything and theres a natural limit which is R0.
They dont trend to anything, they're kept in equilibrium.
Increased transmission doesnt always yield an advantage either. Its energetically more demanding. It might mean more errors. It might be outcompeted by more benign,stable strains.

Lockdowns create selection pressure where only variants more capable of spreading with a thinner pool of hosts can spread.
If that pressure is removed, whatever adaptations MIGHT prove detrimental and come at a cost elsewhere meaning those variants are new replaced.
But that isn't always the case, some have no downside so wont go away.

Lethality is not entwined at all with transmission.
Its transmission window - once thats over the host has no further use.
If extending the transmission window does more damage to the host after that event it doenst matter. A dead host that hasnt been transmitting the virus for weeks provides no evolutionary benefit or detriment at all.

Although viruses often trend to less lethal (this is also due to cross immunity etc) it isnt always the case.
Vaccine induced Polio is a good example where a benign, not very transmissible, non-lethal virus becomes more transmissible and deadly.