Fix error in computing DWS measurements

As concluded during a meeting on implementation of the DWS and TTL effects,

DWS can be seen to measure just the local mosa jitter because although the longitudinal displacement due to wavefront imperfection is significant, it is distributed over a large area, such that the angular error is small. According to https://doi.org/10.1088/1361-6382/aad7f5, a nrad jitter of the distant spacecraft causes a rotation of the wavefront at the receiving spacecraft by 250m, which causes a longitudinal change of <10pm. The corresponding slope of the wavefront is just 10pm/250m40 frad, which is completely negligible compared to the ~nrad jitters we otherwise have. which means that

dws_phi = local_mosa_phi + dws_measurement_noise_rescaled

and similarly for other degrees of freedom.

However, this is what is currently implemented in LISA Instrument

logger.info("Computing inter-spacecraft DWS measurements on TPS")
self.tps_isc_dws_phis = self.distant_mosa_total_jitter_phis
self.tps_isc_dws_etas = self.distant_mosa_total_jitter_etas

This should be fixed to use the local MOSA total jitters.