I wonder if you could test it against a popular dividend ETF like SCHD?

I don't like funds, as a rule. But probably pretty similar, to a first approximation.

The main differences seem to be
* their index is "capped cap weight", cap weight with a cap of 4% for any given company.
* it is large caps only
* they have a weighting on the factor "ratio of cash flow to total debt", which I guess is to ensure resilience through coverage. Not a bad idea, if you're looking for safe/quality companies, which my screen isn't
* they have a greater tilt towards dividend growth rate and long history, rather than dividend now
* their methodology is widely known, so arbs can and will front run it. Not as big a problem as with the S&P 500 or Russell 1000, but probably measurable.
* small fees, and probably some carefully hidden costs too

Presumably because of some mix of the above, the dividend yield seems to be lower. Comparing my estimate of what they actually paid (forward one year payments divided by current price, dates since 2012), to my figures (forward one year indicated yield divided by current price, same date range), their yield came out to 3.45% on the average starting date, the MI screen came out to 4.29%.

Also lower total return if dividends had been invested tax free, difference 2.88%/year. To be fair, their return is a real world result, and mine is mostly a backtest. I was probably overly generous with the trading costs, though. 0.4% round trip is pretty high for large caps these days, especially since it's essentially zero if you use MOC orders.

Jim