Sort:  

Mostly the cons are to us if we don't deliver a useful solution accepted into the code base. In such a case we don't get paid.

The primary potential con to the wider Steem community in such a case is that Steemit wouldn't be simultaneously working on a similar solution, although this would only be a con if we assume they would succeed where we failed. On the other hand, if the solution turns out not to be workable at all, it becomes a pro to the community again, since Steemit wouldn't waste it's time on an infeasible solution.