INR test results should not be considered to be 'accurate.' IIRC, it was the WHO (World Health Organization) (or some other standards body) that concluded that test results +/- 20% of actual (and who knows WHAT actual INR is?), the result should be considered accurate. This is because INR isn't an exact number like, for example, blood sugar. Although INRs taken on more than one CoaguchekXS, for example, may be identical or very close to each other, lab tests on the same blood draw could vary by a few percent.
The point here is that if a person aims for a target of 1.5 - and has 1.5 as a result - if you apply the 20%, this means that the actual INR could be as low as 1.2 -- obviously a dangerous value, whether or not you take aspirin. Aiming at 2.0, the lowest value will probably be 1.6 - although this is too low as far as I'm concerned (and I'm only speaking for myself here).
A person's life will be no different with an INR around 2.5, but the danger of stroke would be reduced.
FWIW - I use automatic, one use lancing devices - 29 gauge, with a pre-set depth. I don't press the device very hard into the skin. I've had no problems with excessive bleeding when using these lancets. (A paper towel pressed against the lanced finger stops the bleeding quickly - usually less than a minute, sometimes a bit longer.
I wonder if you may be setting the depth of the cut too deep or are pressing the lancing device too firmly against your finger. An finger incised for an INR should stop bleeding fairly quickly - especially with the help of a kleenex or paper towel pressed lightly against it.
I've been self testing and self managing for only 15 years, have worked with a lot of lancing devices, and haven't really run into trouble getting the incision to stop bleeding.