the error of the burette is usually ±0.05 ml
Well that depends on the quality of the burette and your technique. The "problem" with a burette is that the marking will be linear whereas the burette's diameter might not be constant.
I think that the burettes I last used were marked at the 0.05 ml mark, but you estimated between the marks. So you recorded to the 0.01 ml amount. It was easy to get at least ±0.02 ml on average for a reading.
I think the ±0.05 ml notion is from 20 drops to the ml. But you can dispense part of a drop by touching the inside of the flask to the tip of the burette.
if I get values such as 14.2 ml, 14.3 ml, 14.2 ml
Then you average and use the values calculated for the mean and the standard deviation of the measurements themselves. It is a "sanity check" to consider the precision of the titration procedure itself.
I suspect that your "problem" is that you're not carrying enough significant figures through your calculations. You only round on the the final results.
So start at 6.72 ml end at 20.92 yields 14.20 ml, not 14.2 ml.
If you get 14.20 ml, 14.20 ml, and 14.20 ml then I'd strongly suspect that your assertion that the precision of the titration is 0.07% is wrong.
The first titration I every did was with potassium permanganate to determine the equivalent weight of an unknown oxalate compound. My titration technique was horrible and I didn't have enough sample left to do more titrations. But I knew few oxalates were soluble, and the sample I had was obviously crystalline. So I did some quick cation tests and found that the cation was NH4+ which has two waters. The prof wanted how I got the right answer to four significant figures with such a poor technique.