Should/do judiciary hide their mistakes?

[NoHtmlMailsPlease]

This is about anglo-society/common law; so don't tell me that
"your particular statutes don't apply".

I read that:
>In the case of Smith v Brummer 1954 (3) SA 352 (O), default
>judgment was refused on a counterclaim because it was held that
>the claim in convention and the claim in reconvention were too
>closely interconnected and there was a danger that a subsequent
>investigation of the claim in convention could disclose that
>the default judgment had been wrongly granted.
------------
I'm well convinced by observation that law-people attempt to
mutually hide their mistakes, as a profession, except when they
are in opposition; but is it normal to *publicly hold* that
justice is to be denied if it "could disclose that the default
judgment had been wrongly granted"? Unusually they just make up
some more peecee/sanatised explanation.
------
And related to 'Courts not wanting to expose a previous wrong
decision', here's my complex matter:-
My pro-se/ed default judgment set-aside application was refused
on the basis of events that happened AFTER the dj.
The appeal judges acknowledged that events AFTER a judgment
cannot justify such judgment, by substituting their own new,
different grounds.
These new grounds were NOT in the record, and I was consequently
denied due process opportunity to also rebut them/it.
An attorney [who knows municipal law well] brought a second
set-aside application, on grounds that 'the claim was void
ab origine'.
That magistrate did not deny the 'voidness'; but as grounds for
refusal, wrote that <such voidness should have been brought at
the first/original rescission application>.

=== My argument is that the second/independant reason was NOT
necessary, because the first application and appeal was sufficient
and should NOT have been refused. But in order to evaluate *THAT*
argument, the Court would have to look at the original appeal.

I fear that a second appeal cannot base their finding on the fact
that the first appeal wrongly concluded that "3+4=6", even if the
second appeal Court can see that "3+4=6" is FALSE.

Ie. that the first appeal's written <grounds for judgment> is
wrong, doesn't matter: "3+4=6" has become a 'fact in law'.
A wrong finding, can only be overturned by a HIGHER
[prohibitevly expensive] Court.

Is that the reasoning in the above case of Smith v Brummer 1954?

== TIA.