My understanding is that it reduced from 30% to 3%. The 5% was a working assumption for the Peregrino valuation report.
What must be remembered is that the original royalty agreement was struck in 2003. Since then Taylor and his lab have done substantial work and have many more patents. Any finished product will involve the intellectual contribution of the original patents with UConn (and the 30% royalty) along with intellectual property that UConn does not have an ownership right to. The difficulty is in determining how much of the end product revenue is attributable to the UConn patents. Rather than go that route, they have instead agreed to a lower overall net royalty rate.
The likely nature of the negotiations with UConn involved the relative contributions of that intellectual property versus the rest of the work done since. The royalty was going to be reduced, the only question was by how much. Better to negotiate this out up front than to litigate it later.