Obviously I'm biased, as I have my own software. But as such I have a pretty good understanding on what some of the issues are whether the more popular software offerings available right now.
I don't know if I followed your walk through fully though, can you break it down into a sequence and not a big wall of text?
1) If you're asking if beersmith calculates the OG correctly based on the users wort volume?
Which is calculated from user inputs on equipment profile, and the user input brewhouse efficiency, then the answer is yes - ish. The cooling factor (adjusting for thermal expansion of wort at boil) is incorrectly applied to the mash gravity, and preboil gravity, OG is unaffected.
2) If you're asking if beersmith calculates the IBU correctly based on the users wort volume, and boil gravity?
The answer again is yes - ish. Beersmith uses a modified version of the Tinseth formula, which itself has only accurate up to +-30% the expected IBU (from Tinseth himself). The two modifications that I'm aware of are
Boil gravity average is used rather than a stepwise of instanteous calculation, so sugar additions added during the boil are not calculated properly.
Whirlpool additions and hop stands are a fudge factor based on a skeptical pro brewery rule of thumb, which does not apply to homebreweries due to the differences in chill times, transfer times. Moreover this does not extend the same fudge factor to additions based before flameout. So if you have a recipe with a 5 minute addition, it calculates to 5 IBU (example), then if you add a whirlpool addition of the same weight it will add another ~5 ibu or so, but the 5 minute addition does not generate any additional IBU which is some sort of magic trick.
Moreover, the tinseth formula was created using his equipment, and process, so your IBUs may not exactly match the IBUs I would given the same parameters. Unfortunately, it's the best we have.
3) If you're asking if the scale function is reliable, then the answer is
Yes - ish. As I listed in your previous topic, there's a bunch of different ways to scaling a recipe. None of them will reproduce the original recipe exactly, each of them have different merits. No matter which way you scale, you will always be changing some aspect of it.
Personally I DO NOT maintain grain %, I treat specialty malts as a spice and keep it at a lb/gallon of wort basis, then scale the base malt to the original OG.
In general, beersmith is the best jack of all trades (currently). But it doesn't do everything perfectly, and is lacking a lot of features that I use all the time so I have to supplement with other software.