Issues in `MarketBidCost` test system data

[GabrielKS]

A few issues in these systems' data that should be fixed soon:

• "c_fixed_market_bid_cost" has ThermalStandards with multi-start costs
• Min gen cost is often set incorrectly

More generally:

• I have not found the "c_market_bid_cost" system super useful in testing, instead I've been creating my own time-variable MBCs with the fixed system as a starting point. Maybe this should just be replaced with those