We have seen a lot of BFs arising from reshardingCriticalSectionTimeoutMillis timing out in test infra due to slower machines. A few examples include BF-30637, BF-30377, and BF-30354. To avoid more such coming up due to users issuing reshardCollection via ShardingTest fixture instead of ReshardingTest fixture, we should increase the default timeout internally inside ShardingTest fixture, while allowing any explicit override.