OT a bit but I guess it depends which approach you want to take. I'm only 12-ish months into my brewing career, I don't brew as often as I like, and I never brew two similar brews in a row, so I don't get much of a chance to develop my own calibrated eye and I've relied on tools/calculators to try to develop consistency. I'm also an engineer and prefer to understand the theory behind what I'm doing rather than developing an empirical "if I add this much grain to this much water and mash/boil for this long I get 1.0XX wort from my system" approach. That's not a criticism of that method, it's just not the one I choose to take.
With this background, I initially relied on the Brewer's Friend recipe builder and brewday logs to attempt to predict my brewing outcomes and calibrate my recipes to my rig. What I found though was that each brew was regularly coming out below the predicted OG and other gravities (in particular the pre-boil gravity) despite re-calibrating with measured data after each brew and lowering the planned efficiency for the next brew each time.
After probably six brews like this I got suspicious and started testing the consistency of the calculators, recipe builder and brew log and was pretty unimpressed with what I found. Aside from the inconsistent and inaccurate efficiency calculations which I have no explanation for, there is also a designed disconnect between the recipe builder and the brewday functions (which wasn't obvious to me but maybe I'm just obtuse) and a bug that seems to be related to deadspace calculations which leads to discrepancies between the two even when accounting for that designed disconnect.
I'm not saying that it's completely detached from reality as the discrepancies are small and plenty of people are happily using it, but there's enough bugs and errors in there that I decided to abandon it and build my own spreadsheet recipe builder using standard brewing calculations and hacking in LC's IBU calculator (thanks
@Lyrebird_Cycles 
). Since doing that, my last six or so brews have been within a point or two of predicted gravities and I now only use BF for the inventory management which is pretty great being cloud-based.
I've certainly made errors in my process at various stages but I'd like to be able to think that a commercial product that is basically a series of simple calculators won't be the culprit when something turns out wrong, unfortunately that hasn't been the case. The other disappointing thing for me is that very few of these tools, including BF, publish the formulae used at each step (they do for some but certainly not all) so it's very difficult to check whether those calculations are functioning as intended. I know that most probably aren't interested in that level of detail but it's when bugs crop up that lack of auditability becomes an issue.
Apologies for the rant - clearly I haven't quite moved on from how pissed-off I was at the time about this and how BF support fobbed me off (while a paid-up subscriber) when I raised these issues despite me providing detailed worked examples of the faults that I'd identified in their system to make it easy for them (I know, I know - I was THAT guy). There were already multiple unresolved threads on their support forums regarding these problems at the time that had just seemingly been abandoned as too hard, and it appears that
@Jack of all biers is experiencing it still. A bit too early on a work day to RDWHAHB unfortunately so I'll just blow-off steam here instead...
TLDR: BF is probably fine for general recipe guidance but there are enough inaccuracies and inconsistencies that I wouldn't recommend it for detailed analysis, calibration, and recipe adaption against your own rig.
