Jack, a thought for future measurements: Bob Thurston, also one of the most experienced certifiers, taught me to record interim points during each measurement. These points could be the preliminary mile or kilometer marks, or just random spots that are easy to mark with chalk or tape, or which can be identified exactly with respect to fixed objects. For the typical 5K, 2 to 4 of these interim points should be sufficient. Record your Counter reading for each spot on both rides.
Retain this data for later use when something like your course change comes up. Using the recorded interim point from the shortest (fewest # of counts) of the 2 measurements in the section that is closest to your area of revision, i.e. a new turn around, you can re-measure just this section with the revised route. When you measure it twice, you again take the shorter of the two, note the # of counts you have added or subtracted (added, in your case), and adjust the start, the finish, or the start/finish accordingly. Of course, if the start and finish are contiguous, you divide the adjustment counts by 2 to lengthen or, as in your case, shorten the S/F line.
Doing it this way prevents you from needing to re-ride the entire course. If the redesign is for longer distances than 5K, you will appreciate having these interim points to go back to. Of course, if the course change is more than a couple of feet or so, you may want to adjust the mile marks. Again, if you have recorded the mile marks on each and every ride, this task becomes relatively simple.
Measuring this way essentially does what Pete's method does. The difference is that you may save some time and effort in future measurements with changes when you already have interim points recorded. Bob Thurston takes this to another step: after the initial two rides, he compares each interim section measurement and then he selects the shortest of the two and enters them in a new column. He then adds all of these shortest sections. Since it happens sometimes that you get a better (shorter) measurement on the first ride than on the second in any given section, you now essentially reduce the course measurement to the combination of the best section measurements.
I am not quite as new at measuring as you, but this "interim" technique is somewhat new to me. I imagine this has been discussed in the past here, before my time on this Bulletin Board. I am curious how many experienced measurers and certifiers regularly employ this technique.