Deprecate & remove "clusterctl alpha topology plan"

[sbueringer]

We recently closed a few issues - which would have been necessary to get "clusterctl alpha topology plan" into a good state - because there was no interest to implement them:

• Support RuntimeSDK in clusterctl alpha topology dryrun: Support RuntimeSDK in clusterctl alpha topology dryrun #6860
• Improving defaulting and validation in clusterctl alpha topology plan: Improving defaulting and validation in clusterctl alpha topology plan #7245
• Make clusterctl alpha topology plan to use SSA / dry run: Make clusterctl alpha topology plan to use SSA / dry run #7246

After a discussion in the Cluster API office hours from 24h January we decided to remove the command.

This issue tracks the removal of the command.

For folks using the command to dryrun the clusterctl topology reconciliation the best known alternative is to setup a management cluster via kind and then run the reconciliation in that cluster (ideally in some way that no actual infrastructure is created, e.g. by using invalid credentials).

While this is more effort, it allows folks to use Runtime Extensions and also solves the other issues in #7245 and #7246.

[sbueringer]

/triage accepted

[fabriziopandini]

Deprecation done.
We can figure out removal or alternative paths, like e.g. descoping the command to generate desired state only; also we can move the execution of this command into a controller (eg. using a new crd) so we can use runtime extensions

[fabriziopandini]

/priority important-longterm

[fabriziopandini]

/remove priority-important-longterm
/priority backlog

We are going to remove this in 1.9
Eventually, an alternative command for computing desired state only could be added with a different name, but this is outside of the scope of this ticket