1. They can still capture data after unblinding.
2. What you are promoting, time in order to achieve a bigger longer tail, may in fact do the opposite. It is the common ( I do it myself from time to time) thinking as if everyone started at the same time. They did not. In fact not only that, the second half not only started later but have less immunity on average. A KM graph will not reflect when they started, it will instead reflect as if everyone started at the same time. Thus if the second half perform less well due to broader inclusion policy, the long tail as it currently is, might be slowly negatively (not positively) impacted over the passage of time. It's a bit beyond diminishing returns. Those censored patients instead become evented patients. Censored helps at this point, the curve does not click downward, eventing does not help the long tail.