chore: deflake multiple invalidations e2e test

This test was failing since **more** than the expected checkpoints were
getting invalidated. This was caused by a slow setup, where more invalid
checkpoints than we expected were created.

Since we only care that more than one checkpoint is invalidated for this
test, this fix just removes the assertion in that at most one extra
checkpoint was being invalidated.

Author: spalladino

yarn-project/end-to-end/src/e2e_epochs/epochs_invalidate_block.parallel.test.ts

@@ -381,10 +381,9 @@ describe('e2e_epochs/epochs_invalidate_block', () => {
       timeoutPromise(test.L2_SLOT_DURATION_IN_S * 4 * 1000).then(() => CheckpointNumber(0)),
     ]);
 
-    // The invalidated checkpoint should be the first one
-    // Note that it may also be a checkpoint *before* the first one that gets mined in `initialSlot + 1n`
+    // The invalidated checkpoint should be the first one,
+    // but it may also be a checkpoint *before* the first one that gets mined in an early slot
     expect(invalidatedCheckpoint).toBeLessThanOrEqual(firstCheckpoint);
-    expect(invalidatedCheckpoint).toBeGreaterThanOrEqual(CheckpointNumber(firstCheckpoint - 1));
 
     // Restore bad nodes back to normal. They should eventually detect that their archive root does not
     // match the value on chain and roll back their invalid nodes.