You'll have a much different tenant base (think collection rates, vacancy rates, damages, etc.) and a much different expense profile. 

Using a combination of assessment data, google street view, count GIS mapping information, and old MLS listings I can usually make a fairly quick screening of each property before going to do a drive-by.

They had next to no data available. 12 lots had homes on them (out of 30 available).

The numbers in the tool are as reliable as the data chosen to perform the analysis, and that is up to you, not the software.The tool currently uses Zillow data to suggest a number of comps for you, however you still have do the work of verifying their suitability for your analysis.

I've also filed a complaint with the BBB and found similar stories there and in two other placesIt seems that they have served some customers well in the past, but in my case they are only collecting money, and not marketing leads.

So for example if you would pay $1,000 per month to rent a similar unit but after you collect rent and pay expenses you are only on the hook for $500 of mortgage payment you would treat that $500 difference as a return and calculate your other metrics from there.

Disclaimer: I'm not a data scientist and these calculations are only representative of my own opinions, take them with a fair amount of salt. 

What I don't like about this approach is the management company would give the job to it's favored contractor, you lose control, cost is inflated, and collectively you all have no control over the cost.

(Looking for indication of calling within 48 hours of initial request, first visit within 3 days [assuming its not urgent],  etc)-When do you initiate collection on past due rent? 

A note about Houston, we lived there for 25 years and still have a house we collect a mortgage on.