.. and I wouldn't underrate manual trim sheets.
As an aside, one needs to keep in mind that the output of the trimsheet is IU and the numeric value of the IU is entirely an artefact (artifact - whichever you prefer) of the sheet's design. That is to say, a sheet designer can make the IU whatever sort of values he might choose for whatever strange reason motivates the design philosophy...
If the sheet is well conditioned in its design and executed with a modicum of care by the user, the calculated CG can easily be within 1-2 mm of the longhand calculation results. However, many trimsheets are designed with little consideration for error management and the results can be highly variable.
I can recall a sheet I designed many years ago for a corporate operator. The semi-graphical/longhand system which they had been using previously was a bit of a pain and they wanted to simplify the calculation exercise.
As the particular aircraft presented some design difficulties for the trimsheet (one of the reasons it hadn't been done before) I put a lot of effort into the job and we ended up with a very nice product. (The sheet design subsequently was pinched by all and sundry and I would not be surprised to see it surface on this Type in all sorts of strange places ..)
Outcome was that the first asy training session using the trimsheet with their previous standard training ballast configuration showed the CG to be out of limits. They were a little sheepish when I investigated and showed that the previous system, due to its poor design conditioning, was sufficiently error prone that they had been conducting such training routinely outside the envelope (not significantly, but it could have been embarrassing at the the Enquiry).