ci, qa: Adjust timeouts per test instead of global `--timeout-factor` #23476

A timeout failure in a functional test could point at a flaw in the test itself, e.g., #23463 (comment)).

Therefore, setting --timeout-factor globally can just hide such flaws.

This PR drops non-default --timeout-factor in CI, and adjusts timeouts in three tests only.

Can you explain how a large timeout can hide something? If there is a bug that will result in a long timeout, it surely will result in a short timeout just the same?

This will likely break the valgrind config, which is not run by CI here,

@MarcoFalke

This will likely break the valgrind config, which is not run by CI here,

The TEST_RUNNER_TIMEOUT_FACTOR variable can be assigned with a proper value in 00_setup_env_native_valgrind.sh, no?

Can you explain how a large timeout can hide something?

From your comment:

test_feefilter generates blocks and txs on node1 (without syncing with node0). Then node1 won't accept/request the txs (because it is still busy syncing the blocks). Thus the test fails (times out).

If node1 incorrectly rejects the tx, the test will time out regardless of the timeout factor. Increasing the factor can at best unhide something, not hide something.