As an observer, one should never assume a patent is not going to be granted in one territory because it failed in another.  Further, one should never assume that something very basic will not meet patentability criteria somwhere and dismiss a published patent claim as ‘unpatentable’.  We see companies placing too much reliance on this in the generics industry, at great expense.

As an IP generator, I would suggest filing an application for every little modification you make to the prior art, and if that’s not cost effective publish it on IP.com or similar to stop someone else from patenting it.  The patent quality issues won’t change overnight, so if you can’t beat them, join them…

Finally, never stop watching the patents and updating searches. Too many companies fail to realise that it takes 18 months from priority filing to publication. As you say, a slight change in the subject matter when working the invention could make a narrow search obselete – so try to search a little broader intially – it may cost more at the early stages but strategically you know where you can move (and where you cannot) if an identified patent is granted…