This works in some markets, but it others the variability makes it useless.

My reasoning is that there are now too many variables we cannot control.

For example, a HELOC is a variable interest rate...so it's hard to know what your payment will be in the future. 

Hopefully I can sell as a package or within the 1031 timelines allowed, which is definitely a variable of risk, but I have a couple investors I've turned down over the years who may be interested.  

I agree with Brian Ploszay “An algorithm doesn't have the correct inputs to make a decent valuation. “ the algorithm does not take into account the “inside” of the property, nor does it know of conditions in the local market or other fluid variables.

As I mentioned, get specific quotes on some of the items to nail down the more variable pieces.    

My ROI: Depending on what variables you include, ROI can become confusing.

Reason for this as follows: You control the input and make sure it's real, you can also explore scenarios to see if the deal work, i.e. 220k sales price doesn't work, what if we sold at 230k, would that work (make sure to check market will support you explored changed variables)?

A HELOC rate is variable and is usually higher then a 30 yr fixed.

I looked into other projects recently built by a developer we would like to recommend and determined they typically complete projects of similar size and type inside of two years.I understand many variables can affect both demolition and construction time, but I am only looking for a general rule I could apply in the absence of reliably sourced data.Thanks for your time!