Although many aspects are good, there is one potential problem with some of the suggestions about RWSL, particularly if it is only installed at an airport where it is perceived that 'there is a need'. The mindset with RWSL is that when the red lights are off, we can cross/roll, when you then arrive at an intersection or line up on a runway where there is no need, at least half the protection of the system is lost. There is an additional gotcha if you are used to operating from airports with RWSL and then find yourself somewhere with a less sophisticated system, the usual mindset can be a problem.
I've also got a slight hesitation about the control systems, which are driven by ASMGCS data, and rules about when the RWSL is activated - having had some involvement in setting up RIMCAS (Runway Incursion Monitoring and Alerting System) at a pretty simple airport, getting the rules correct so there are no false alerts is not easy. With RIMCAS, the alerts are 'only' raised to ATC and humans can assess whether the alert is legitimate, although there is often little time for this (or, worse, an alert is ignored because 'it always goes off when this happens'), RWSL, as I understand it, is fully autonomous - the system is most useful at complex airports where the ruleset which does not produce erroneous operation will be difficult to compile. And then, of course, one day ATC will say, "Ignore the THLs, clear for take-off". The tech-savvy will now start talking about AI......