DAL: increase baker timeout

What

Increase the timeout the baker waits for the DAL node to answer to the GET attestable_slots query from 0.5s to 0.75s.

This was actually already sneakily done, for the same reason (see below), in !19734 (merged), but only for Alpha. This MR now does it for S and T as well.

The MR also contains:

• two small commits concerning the RPC_server.warn_missing_shards function
• adding an information about the level in a baker event
• a commit that slightly modifies a test such that after each bake_for we wait for the DAL node to process the new block

Why

The reason to increase this timeout is to make the test S023->T024: test skip-list RPCs with migration fail less often in the CI.

Note that the only reason to have this timeout is for tests that (usually) bake in the past, so there is no time between blocks. Without this timeout there would be no time for the baker to obtain the required DAL attestable status of its assigned shards.

Also, I have observed no failing runs when the function RPC_server.warn_missing_shards is not called by get_attestable_slots (and without changing the timeouts)... But it's nice to have these warnings, so I think it's better to increase the timeout.

The tezt commit was the first attempt to make the test fail less often, but it didn't help. I think it also doesn't hurt, so I left it in. (BTW, we could do this waiting directly in bake_for_and_wait_level; to be tried/discussed at some point.)

Manually testing the MR

tezt -t "S023->T024: test skip-list RPCs with migration" --loop-count 20 -j 5