This question ought to generate some discussion.
One should review the basics first -
(a) the rules in a particular jurisdiction are what the appellate court(s) judge(s) determine(s) them to be after the inquiry into the inquiry (ad infinitum) into the crash.
(b) apart from (a) the purported rules for a given operation will require the consideration of matters such as
(i) the plan area to be considered for obstacle clearance. This normally is some sort of quadrilateral-derived shape and is superimposed over the intended aircraft flight path. The actual shape varies from Standard to Standard.
If a lumpy bit is inside the area, it is to be considered, if outside, it WILL be ignored.
As a result, the pilot MUST track the nominated escape path (within the nominated speed and bank angle range) with critical precision, as some runways may have quite extreme obstacle hazards IMMEDIATELY outside the (rather narrow) plan area to be considered - you go outside - you might very likely die. This is especially the case where the ops eng guys call for turns to avoid terrain.
(ii) the calculated net flight path (ie expected OEI performance less a specified fudge factor) is then compared to each obstacle height plus a small margin and, if it is above the obstacle height so calculated, the obstacle can be ignored. Otherwise either the RTOW reduces or the ops eng guy looks for a different escape path.
So, to your question -
(c) Each source of data will have a inherent and able to be determined accuracy probability. There are various approaches to addressing such errors in ops eng calculations
(d) The error probability will vary according to the means with which the data has been determined, especially if photographic based.
(e) The error probability then, itself, increases with time from the date of data acquisition involved in the production of the final data. For instance, a map can only be as accurate as the original aerials and any subsequent field checks used in its construction, and then only up until the date(s) of those activities
(f) maps generally give ground heights, not tree heights, for instance... and the vegetation changes constantly.
.. the problems just go on and on ...
As an ops engineer, my attitude is
(g) I do the best I can considering the time and data available
(h) I apply additional conservatism if, and as, I consider it appropriate
(i) if I have time and the need is appropriate, I will consider a specific survey of critical obstacles
(j) I endeavour to have the operator's flight standards organisation institute an ongoing flight deck monitoring of the general area in the escape path
What does that mean in terms of your question and my original point (a) ?
I suspect that no matter what you or I do, the lawyers will always try to find fault with it.
It follows that NO level of error is acceptable.
It is a matter of deciding what you, as an operator, pilot, or engineer, can live with in an environment of inaccuracy. Subsequently you can be called on to justify your decision(s) and be liable to penalty if the legal system decides that your decision was not quite up to the mark.