Explicitly model resource limitations

[denised]

In some cases, integrations adjust adoptions of models to reflect the competition for common resources between them. This has the effect of muddying the derivation of the adoption, and also typically only impacts a given scenario. Instead we could have an explicit model for the availability of common resources, and the result of integration would be to modify that allocation. (There would be different allocations for the different PDS1, PDS2, PDS3 cases). This would have a couple of benefits: it would make it easier to make consistent scenarios, and it would be easier to experiment with different allocations.