The other answer is that having a sudden break that you don't want interrupts what is called "flow" - this is a state you enter into when fully immersed in something, see http://en.wikipedia.org/wiki/Flow_%28psychology%29.
One bar of silence or more for a pick up obviously makes sense in some contexts, particularly when you have a pick up such as the vocals you mention before. Vocals also have a clearly define linear progression; it is important the the words follow in a specific order. So I used this technique (in Edison) just last week to record 3 verses one after the other by looping the verse with two bars playing before the start of each verse because the first words fall prior to the first bar of the verse. So this method makes perfect sense in the right contexts (and is by no means new to me). I already knew exactly how long each verse was - because I'd written the lyrics to them and they were only so long.
But there are other cases where I simply want to be able to jam for 2-5 minutes over the same loop.
Use Case 1: Composition from a blank slate
When: Writing a new song
Aim: Come up with the basic building blocks of a song
I'll typically lay down a basic 4 bar beat and will then jam over it on guitar/bass (or keys but that's MIDI and not the issue). I want this loop to keep playing while I jam, and try out different ideas while recording them. A sudden dropout or a bar of silence completely destroys flow, ripping you out of the creative process. As a musician, surely you understand the importance of groove? It might take a while to find the groove you're looking for; you don't get into the groove if you're contantly stopping and starting. And when jotting down rough ideas, it doesn't matter if your loops aren't completely clean with notes hanging over from the previous loop (if that's what you're concerned about). This can normally be fixed/mitigated with a minor bit of editing and anyway, these aren't ever going to be the final takes. Instead I'll go through my 3 minutes looking for those takes where I like the groove, and take out a couple of those. Then I'll record the next instrument in the same manner. Once the basic melodic groove is in place (guitars, bass, possibly vocal placeholders "la la la") I'll go back and rework the drum beat and all other MIDI tracks so that everything fits the groove I've come up with. Then when I have my beat and other elements in place and everything grooving nicely, I'll re-record each part properly and cleanly (and normally linearly!).
Use Case 2: Tape is always running
When: A bunch of people are jamming
Aim: Record ideas as they occur for later review
Especially when jamming with other people, you'll often just want to hit record and jam out - you have no idea how long you're going to jam for, and you might as well record everything to review later. Hard disks are cheap, the Del key is easy to access, so who cares if you don't record anything worthwhile. Set up a basic beat, repeat it ad inifinitum and jam over it to your heart's content. If you come up with some great ideas, you have them captured and can use them as a basis. Again, having a bar of silence won't work and will destroy any semblance of creativity; audio dropouts both disrupt the flow and mean the recordings are worthless. The only alternative is to repeat the beat 29892489233 times in the playlist.
Use Case 3: I'm a poor guitar soloist
When: I need a guitar solo
Aim: Hide the fact that I can't solo
I've never been a lead guitarist and don't get off on sitting at home alone playing back other people's solos on YouTube. So I'm primarily a rhythm guitarist. I can kinda fudge solos, but I can in no way simply hit record and play a brilliant solo in one go. Instead, I'll jam over the instrumental part for a long time, and then will take the best parts from the various takes and put them together. Having a bar of silence again ruins the flow - and again, if the instrumental part ends up being 16 bars, a bar of silence every 5 bars is a no go. Also, recording a guitar solo part means you don't normally have to worry about finding suitable places to cut the take, as there are natural pauses (unless you believe a guitar solo is about hitting as many notes as you can in as little time as possible). In this case I'll normally use the parts I record in a loop as my final takes too. Sure, it's a crux, but isn't that why it's there? To make things easier?
Use Case 4: Repetition is everywhere in music
When: A part of the song consists of the same element repeated multiple times, e.g. a verse based around several repetitions of the same progression
Aim: Get a nice take that lasts several times the length of the loop, and the whole length of the part that's repeated multiple times
A verse often consists of 2-8 repetitions of the same background. I ideally want a good take that's exactly as long as the part I'm recording. But what if I screw up in the first loop? By your rationale, I'd wait 7 repetitions, then count in a bar, then retry. That's a time waster. I can just as well start again from the beginning with the second repetition, and keep the take lasting until the start of the ninth repetition. Who cares if my final take is loops 2-9 not 1-8? But I can't do this if there's a bar of silence or audio dropouts part way through.
What's common to all these use cases is that I'm not looking to record four bars. I'm looking to repeat four bars. I may want to record 8 bars, 16 bars, 32 bars. I may not know how many bars I'm going to record, particularly in the case of a solo.
Now I'm definitely not the only person in the world who likes to record my ideas as I work. Maybe it's just restricted to the people I know personally. Maybe we're the only people who think that stopping/restarting can be a massive workflow killer, and that it's much easier to deliver a good performance when you're in the moment and only concentrating on that. This has nothing to do with an inability to count a bar of rests (strange that you'd think this is the most likely thing to be tripping people up). Now I do know that a lot of classically trained musicians can't work like this - tell them to "jam in A" and they are completely out of their depth. So maybe that's why you've never worked this way. I've worked with loads of musicians for whom this approach was perfectly natural: get in a sax player, tell him to groove away for 5 minutes and then chop and edit the results to taste. The sax player's not like "woah I need a bar of rest, I can't jam for 5 minutes". Tell a guitarist to solo, and he might wank away for 2 minutes, delivering a perfect solo. You may have no idea beforehand how long he's going to solo for; the last thing you want is for a sudden one bar break to appear just as he's reaching his climax, throwing him out of his flow and destroying the solo.
-----
I know the workarounds, so don't tell me to loop 16 bars if I want to record 16 bars. As I said, I may not yet know how long the verse or solo is going to be, because I'm in the process of writing the song. Getting a studio musician to come in and record a part over an otherwise complete piece is a completely different kettle of fish.
-------
Also I don't see why implementing HD recording is such a big deal. The technology has been around for 20 years. I don't see how FL's inability to do what any other DAW can do (and have done since day 1) should be dismissed with "you're doing it wrong" when the implementation is clearly what's wrong. There shouldn't be dropouts - whatever your workflow. How are dropouts really acceptable? Is this really just because you can't be bothered to argue with gol again?
What daw products do you actually own, as in "have actually purchased"?