John, no it doesn't quite cover it (although it should cover most of Charlie's question). I envisage a situation which is common to all LRCs, where we are doing a datasearch let us say for all records within 1km of a planning application site. The process you describe will return all records where the spatial reference falls within or clips our search polygon (a 1km buffer around the site). That means it will also return all records associated with a location which has a spatial reference which falls within the search polygon, where those records have no unique spatial reference but use that inherited from the parent location.
However there is a third class of records which should be returned and which as far as I see are not. Records of taxa associated with locations, the boundary polygons of which intersect the seach polygon, but where the spatial ref of the parent site is actually outside the search polygon will not be returned.
It is my understanding that when occurences are recorded against a site (location), but no additional spatial reference is supplied (I personally have collected in excess of 50,000 such records in the last five years of my employment), it is implied that those ocurrences may lie anywhere within the area defined as that site. They may be widespread or ubiquitous within the site, or they may be rare but lacking any more detailed spatial information. Either way if we ask for all occurences within 1km of a development site, we should receive all such occurrences which may fall within the area where the search polygon intersects the site boundary polygon.
If Recorder was able to return such data we would be able to use it for all area-based data searches in one single query. Of course it would be useful to recieve also an indication that these records are rather less certain than those which we know were definately recorded within the search area, but that is mere refinement.
It may be that I have missed something in the documentation and that Recorder actually does precisely what I describe, but this has not been my understanding so far.
Rob Large
Wildlife Sites Officer
Wiltshire & Swindon Biological Records Centre