I picked up at least a few minutes per section when I started bubbling a passage, game or page at a time. The few seconds it takes to bubble after every single question really adds up.
It would take, conservatively, 3 seconds to bubble after each question. You mean to tell me that 1.5 min matters?
It's not even that much--- IMO it's maybe 1 second to bubble and 1 second to "reset" back to the page, but lets run with 3 seconds.
On a 25 question section that's 75 seconds out of 2100-- put another way it's ~3.6% of your time for the section which does sound like a lot.
But you have to remember that "group bubbling" isn't free either--- figure .25 seconds to write down you answer the first time (I'm assuming that you have to write the answer down on the test as a placeholder. I used to do SAT bubbles 4 or 5 in my head and transfer them all at once, but I feel like that's much more difficult with the LSAT, which IMO demands a higher level of mental engagement), and figure .75 seconds per answer to move them over as a group-- and these numbers are again extremely conservative towards single bubbling / generous towards group bubbling, I think.
That leaves you with 25 seconds for a 25 question passage. Which means that the difference between the two methods is by this estimate 50 seconds, or ~2.4% of your total time for the passage. Again, the percentage sounds like a lot to my ears.... but 50 seconds, not so much-- and that's at the *high* end (I'm assuming that you have the answer sheet and the question book side by side like for the SAT--- if the answer sheet is fixed to the back of the booklet or something scratch everything I've said and group bubble).
Personally I know that I had misbubbling mishaps a few times on the PSATs when I was using the group method-- though I always caught 'em (but that came at the sacrifice of time of course). I've done in total maybe 10 practice tests in the last 8 months, all using the group bubble method. I've misbubbled, to my memory, one question in that entire time, and I'm sort of a space cadet. (I don't skip questions ever except rarely on games, which does help a lot. Though skipped questions probably make both methods more error prone).
The accuracy advantage of the single bubble method is huge, IMO, and the consequences of a misbubble are largely mitigated.
Sure, if you're confident and you feel that a few seconds can make the difference for you, there're some reasons to groupbubble.
But singlebubbling is pretty frigging quick, and for me at least I've found it to by ~99.9% accurate. That's where I'd put my money.
JMO though, and in way more detail than was necessary, I'm sure.