No, this is absolutely wrong.
You MUST understand the limitations of precision of measurements in order to understand your results. Your answers will NOT be more precise if you merely use more digits. You need more
meaningful digits.
And when you do calculations based on many numbers, the errors add up. And add up far faster & to far higher errors than you can imagine.
But these errors are REAL.
A classic example is doing a "tolerance stack up" in an assembly. Even tho all the individual parts might be made with high precision, after you add up all the tolerance errors in all the components, you find out that you can't guarantee the location of some component, except within some large error.
So you need to take other measures...
In the days before everything was CNC machined, this was absolutely critical. You could tell the quality of the designer merely by looking at the tolerances on his designs.
Crappy designers (or kids) put ±0.001" tolerance on everything.
Good designers put that sort of accuracy on only the parts that required it, and looser tolerances on everything else.
You MUST recognize the total error in any calculation that results from the errors in the several measurements that make up the variables.
When I was getting my engineering degree (in 1974), if you handed in an engineering lab report without an "error analysis", you got an automatic F grade. The professor wouldn't even bother to look at the rest of it. THAT is how important awareness of the errors in experimental analysis was considered.
When you use more digits that are warranted by the measurement technique, you're simply lying to yourself about the accuracy of your answer.
There were several issues with trying to digitize data off of an tiff image of an old vellum drawing. The most important is resolution of the image scan, the fact that the vellum will change size over time, and (most important) the fact that draftsmen are careful about where they put their lines ... but not THAT careful.
Everything after the 3rd digit in his analysis was merely random numbers, which had zero relevance to reality. His 10 digit numbers were LYING to him & to any (innumerate) person who believed his analysis.
THIS was a consistent patter with his lack of understanding of the real meaning of numerical analysis. He had no feel for numbers, and it led him to making ludicrous claims about his accuracy & precision.
If you do not understand the INaccuracy of any measurement or calculation, then you have NO real feel for the number.
Listen to a beloved, renown expert:
https://www.youtube.com/watch?v=md8G_cTHa60&t=4m30s
“All important in making measurements … is the uncertainty in your measurement. Any measurement that you make, without knowledge of the uncertainty, is completely MEANINGLESS.!”
__
This is one of the serious steps backwards that happened to engineers when we switched from slide rules** (which is what I had to use when I started) to calculators.
When you use slide rules, your best estimates were limited by the accuracy of your placement of the sliding central bar & the cursor. For those that have never used one, the transparent cursors was a slight magnifying glass to help. But 3 digits of accuracy was all that you got from your calculations. And, frequently, 2, 3 or (at best) 4 digits of accuracy was the best you could get out of your measurement instruments, too.
** Please guys, I've heard several people (writers, TV, Movies, etc) refer to them as "slide rulers".
NO.!! BAD faux-geek.
But, as ever, 'it depends'.
