Except the data from RIO doesn’t reflect that whatsoever.
There have been 1,902,988 +10 keys completed and 1,252,841 of those were timed. Yes, this doesn’t include the failed runs, but it shows that a large number of the completed runs are timed at a rate of 65.8%. The failed runs are not going to drop this number down to sub-5%.
I don’t understand where you are getting the assumption or “data” that <5% of attempted +10s are being timed.