Before going further, I just want to confirm that you are not entering any (or hardly any) start or finish dates and that you are letting project calculate those based on predecessor/successor dependencies and durations.
There are several reasons why your calculation isn't a good idea.
First, unless the task is on the critical path the finish date is purely an academic exercise. In fact, the task's real finish date is the last date which won't make it (or a dependent task) late that, in turn, will put those tasks on the critical path or push the critical path finish date out. That 'real finish date' is the "Late Finish" date in Project.
For corresponding reasons, the Start date is also a somewhat academic exercise because the task doesn't actually need to start until the last date which won't make it (or a dependent task) late that ... well, you know from the above para what comes next.
Second, you need to deal with tasks which are in the form of n-days of work over n+m days duration (5 days of work over 9 days duration, for example). For ease of explanation, suppose you have a task with 1 day of work to be done over 10 days. Suppose 0 days of work have been done during the first 8 days. Is this task late? What if it's on the critical path?
Your calculation assumes that the estimates of work and duration are correct. They rarely are. That's why I *never* (unless I slip) ask resources what percentage of the work they have completed (old joke: the first 90% of the work takes the first 90% of the time, the remaining 10% of the work takes the remaining 90% of the time). I ask how much work remains and how long it will take to do it.
Then there are the technical reasons which will cause some difficulty in performing the calculation. If the project calendar has a statutory holiday in the middle of the task then the equation needs to reduce the number of working days available. If two resources are assigned to the task and one of them has some vacation scheduled then the equation needs to factor that into the equation.
There are other issues, but these are enough to get you started.
Me? I focus on tasks that are on the critical path (identified easily enough) and I also look at tasks that have (and the number varies here depending on circumstances) 2 (or fewer) days of slack.
I said I'd give you a calculation based on your formula. I'll work on that later today.