I believe that the problem here is Southern's cost-benefit analysis. I've seen arguments in my industry which justify automation in terms of labour savings.
At first glance it would seem much cheaper (and more sensible) to install Oyster Card readers (and gates) with access via the car park at Perry Vale instead of installing two lifts (at a cost of ?200,000). This would enable easy disabled access to the down Platform, encourage car users to park in the car park rather than on Devonshire Road and maximise revenue streams. However, if these gates have to be manned, then the labour cost needs to be added in; there is no consideration of cost/benefit to their customers