I did a programming test for where I currently work. The biggest difference is that it would not have worked for many realistic use cases, and the code was fairly trivial.

If you have to implement an entire application or system, that should be an automatic red flag.

That's the differentiator; I have no problem with that.

> Don't do free projects for companies. It's never worth it.

In my experience, I'd never get hired if I took this approach.

Work samples are quickly becoming the interview standard for engineering. In fact, I find they're often completely replacing all day onsite interviews.

I'd much rather do 8 hours of self-directed, self-paced work than get grilled on a whiteboard during an on-site.

Let me be clear: I'm okay with NON-business specific challenges. AKA, from a marketing perspective:

How would you position a new SaaS service that allows you to track your children's school performance? (but the hiring company is in SaaS business tools)

Vs. doing work that is directly actionable and can be taken by the business, "How would you update our homepage?"

That's the line now. Not opposed to challenges.

The company I work at pays a lowish rate for our 10 hour take home project, which is also the last step of the interview process. I think this is a great practice, as this shows that we're taking this seriously and are willing to compensate for your time.

Does your company then issue 1099 forms (assuming US taxation)?

They might not have to if it's a "lowish" (quoting GP) enough rate. I believe the threshold is $500-600 per annum, so if they're paying you $50 per hour for the interview, no 1099 needed.

$600 is the cutoff for requiring a tax form to be issued.

Perhaps be careful - we did this until someone filed for unemployment benefits. They used the fact that we paid them as evidence of employment to the state.