So I was writing up my referee report and wanted to encourage the author(s) of the paper I was reviewing to use leading zeros. I sought for support from the Chicago Manual of Style, 14th Edition. It said:
8.19 In scientific contexts decimal fractions of less than 1.00 are set with an initial zero if the quantity expressed is capable of equaling or exceeding 1.00.
8.20 If the quantity never exceeds 1.00, as in probabilities, levels of significance, correlation coefficients, factor loadings, and so forth, no zero is used.
where I have omitted the examples the Manual provides.
I had never heard of this surprising distinction before, having erroneously imagined that there were only two views on leading zeros: "yes" and "no".
I will confess that the Manual's take seems a bit too much like Fizbin to me, but I am surely a better person for having learned about it.
Oh, and I suggested that the authors of the paper I was reviewing use leading zeros all the time.