All good points.  I think we both come to the same conclusion: context (or terrain) is everything.

I am not sure I see the issue as one of when to settle, but of when to sue or whe to be ready to sue.  This should be handled by the IP owner as any other market analysis–when is the right time to enter the IP enforcement market? 

In some circumstances, it may be appropriate to sit back and let the market, and attendant infringement, develop.  Moving too early can result in the putative infringer changing its product to avoid infringement, thus leading to lower potential infringement damages and, as a result, smaller settlements.  In the States, an IP owner can sit on the sidelines and let damages acrue for up to 6 years with no downside.  Thus, the IP Owner can watch the market develop and damage acrue, thereby making it more financially lucrative to bring suit (and possibly force an earlier settlement). 

Other circumstances could point to engaging with the putative infringer(s) early to get them to choose the IP owner’s technology for use.  This scenario presents itself when there is an emerging market, and technology pathways are still emerging.  Others may wish to acquire access to IP to provide them a proprietary interest where they otherwise may not have one. 

Certainly other scenarios exist in addition to the 2 set forth above. 

The key to successful decisions on whether/when to sue or engaging in licensing talks with another entity is understanding the market forces and economic factors that exist in the relevant technology space.  All too often, IP owners look at their patents as a legal right that needs to be avenged, as opposed to an asset must be fully understood before being suitably exploited.   I think that IP owners, whether in the software space or otherwise, would be well-served by stepping back from the IP issues (and leave their lawyers out of the picture–at least initially), to engage in a market-based analysis prior to making any decision on how to extract value from their IP rights.