From the perspective of someone who's dealt with finance and exports most of their life, CSV files of this nature are more of a gentleman's agreement between engineers (who all died years ago) than a statement of fact. I don't know how you'd convince a judge that your csv file is error-free when the transaction process itself is full of clawbacks due to all the buggy data and creaky code.
In practice, you'll have created a billion-dollar SaaS corporation and helped write legislation. The fundamental issue is that data quality and handling practices mean this information will always be "truthy". At a base level your input comes from an office clerk who's medium-engaged in the process and who's asked for quantity over quality.
The entire financial network (imho, from my perspective, etc etc) is "best effort" and mutual agreements. Nobody involved wants to cast a string to a legal agreement.
Liens can be challenged too, but there's a presumption of validity that increases with age, and property sales are hard to do without clearing liens, and challenging a lien seems difficult and time consuming, so it is ignored until a sale is needed with some urgency, chances are the lien will be paid rather than challenged.
I don't have any specific thoughts about wage garnisment, but the % of people that don't get on-the-books wages in the general population may be significantly different from the % in the population that has become the target of debt collection.
Anyway, how many of the debts can be turned into a lien or a garnisment is a question too. You've got to turn the line in the csv into someone you can plausibly sue; and you've got to have upfront capital for the court costs (which probably you can add to the lien, but maybe not?).