Pocket nodes, as per the current economic design, are really just market share points (as @RawthiL has helpfully described elsewhere). Having more of them increases your overall share of the relay pie, but they don’t cost much to run, other than the stake amount, so it’s the underlying blockchain nodes where the real costs lie. Answering this question will be crucial for the economics R&D workstream for v1. It’s important that we understand what we are optimising for across the system.
You can play around with this network costs spreadsheet put together by @shane in the meantime - GoP (Game of Providers) - Google Sheets - and check out the full thread here
Remember that most of the supply side runs nodes non-custodially using POKT held by others. If node runners leave, those POKT holders can move their stake to another node runner. Alternatively, the reduced inflation of c.4.98% inflation proposed by ARR means that there isn’t as much of an opportunity cost if POKT holders prefer to just hold and wait to see what to do next. It shouldn’t be a binary situation requiring stakers to either stake or sell.
We will likely make a proposal to the DAO in August about a liquidity incentives plan for wPOKT, so moving POKT tokens across the bridge to stake as a liquidity provider in any new wPOKT pools will definitely be a new form of utility for POKT!