Humans work in base ten almost universally so making calculations in base ten seems quite reasonable. 

The exactness of the results of DFP calculations makes this a perfect number format for commercial or financial work. 

Many, if not most, databases are using decimal storage for monetary values, often as character strings in base ten, so that the possible application of DFP to these databases is obvious.

It would probably also result in smaller storage requirements and would certainly result in more accuracy or fewer checks for edge cases that cannot be allowed in financial calculations. 

The logistic example above clearly shows that DFP results are sometimes “truer” than binary floating-point. 

The preservation of knowledge that there are meaningful trailing zeros in a value is also potentially of significance. 

The Python example returning “0.10” as the answer instead of “0.1” is helpful as it tells users that the digit in the hundredths place is truly zero and not simply unknown. 

A scientific result might require a great deal of effort to know that that digit is exactly zero so it would be good if the computer representation preserved it.

The thought perhaps should not be “should I use decimal floating-point?” but rather, “Is there any valid reason why I should not use decimal floating-point?”

create table test.tbl (c1 decfloat(34), c2 decimal(31,15));

insert into test.tbl (c1, c2) values (1.003, 1.003);

select * from test.TBL;