fix: adjust_pool_bounds fails for more than 1 protein

[BenjaSanchez]

trafficstars

As pointed out in https://github.com/SysBioChalmers/GECKO/pull/28, adjust_pool_bounds currently fails if it's used with 2 or more essential proteins = 0 (the resulting model does not grow). Fixing this is key for geckopy to work with any proteomic dataset and always create a functional model.