There is no simple answer. Some bullets manage OK, some not so well, some fail entirely, and I've never seen a guide as to which models do and which don't. That's other than the 'boat-tail angle rule', anyway. Bryan Litz says the ideal boat-tail angle is 7-9 deg. Go much above 10 and it's too steep for the air to follow the bullet sides around to the base. This seems to manifest itself as much increased drag and turbulence leading to instability in transonic flight.
It's this effect that has led to the common advice of don't use 30-cal 168s at 1,000yd. That is misleading advice as it came out of the 168gn Sierra 'International' aka MatchKing with its 13-deg BT angle, a 300-metre specialised design originally and the various near copies on the market from Speer, Hornady and Nosler. Berger 168s are designed as long-range bullets with 8.9, 8.5 and a really nice 7-deg angle on the BT, VLD and Hybrid respectively. Hornady A-Max 30s other than the 208 fall into this enforced shorter range bracket too thanks to their BTs. The 155 is a massive 13.5-deg, 168 12.8-deg, and 178 12.6-deg.
Even this 'rule' doesn't always seem to apply. Many older L-R Service Rifle shooters talk about good results at 1,000 with some batches of 7.62mm match ammo in their 20-inch barrel M14s using the 168gn SMK.
You can tell it is a transonic issue as it's also velocity dependent, and we usually think '308 Winchester / 7.62mm military' here. I've successfully used Hornady and Sierra 168s at 1,000yd in 30-cal magnums which are driving them fast enough to just keep out of trouble at this distance. Still not recommended of course thanks to their low BCs compared to better long-range speciality bullets.
I was much exercised by this worry in the early days of F Class, shooting a .223 and 80-grainers at 2,800 fps MV or even a bit less. Even the optimistic G1 ballistic charts of the time said they'd be subsonic at 1,000. (Bryan Litz's Point Mass Ballistic Solver 2.0's program says 1,078 fps at 1,000yd at 2,800 fps MV in standard conditions for the SMK; below 1.2 MACH beyond a point somewhere around 780 yards.) In fact they shot fine in a large range of conditions apart from needing around 60% more windage allowance than the 6.5s that were appearing. The biggest problem apart from my wind-reading skills was constantly getting out of the rhythm to call to have the target pulled as the butts crew didn't hear the subsonic bullets and had trouble seeing their little holes.
In the early days of F/TR I used a 24" barrel factory tactical rifle that was billed as F/TR ready - it wasn't! The much touted 175gn Sierra MK as used in the US military M118LR sniper round was allegedly good at 1,000yd at 308 velocities - it wasn't! It would group OK in stable conditions, but any significant change would cause a much greater deflection on the target than the ballistic charts predicted, so transonic flight was obviously making it barely stable. I also suspect conditions on the day had a big effect as Litz's program says it's just subsonic at 2,650 fps MV at 1,000 in standard conditions. Throw in MV spread and there was a risk of some being supersonic, some not and get warmer / colder air moving onto the range under some conditions might change things for the better / worse. I used the combination on Scotland's notorious Blair Atholl range at 1,000 in one competition in a day of cold headwinds from the north and frequent rain squalls. The temperatures plummeted during the squalls (and the wind went mad too!) and what was an 'interesting group pattern' outside of squall conditions changed to seeing me do well to just stay on the target frame at all - and they are BIG frames at Blair!
On ranges other than Blair (which is electronic, so no butts crew), target markers reported they heard faint supersonic 'crack' and saw round holes on the paper, so they appeared to remain stable and just supersonic in summer shooting conditions. Confirmation of this transonic performance phenomenon has since come from USMC snipers who say the M118LR's performance 'falls off a cliff' beyond 800 metres (900yd), which is just what I found with the bullet at slightly higher than M118LR MVs. A move to the 190gn Sierra MK with Viht N550 keeping the MVs reasonable gave a vast improvement in 1,000yd performance.
Incidentally, the old long-range 30-cal SMKs with their extra length boat-tail sections have a superb reputation for stable transonic / subsonic flight. They were used by GB and British Commonwealth 'Match Rifle' shooters (.308W shot at 1,000, 1,100 and 1,200yd) for many years before the current bunch of 210gn and up VLDs and Hybrids appeared - that's the venerable 190, 200 and 220gn MK bullets.
The 'easy' / better answer to all this is to use a design such as the 30-cal 185gn Berger LRBT with a reputation for good long range performance and to load it to achieve or exceed 1,350 fps at 1,000yd. If I can get the combination I'm using to be predicted to hold 1,400 fps at this range in a G7 based program calculation, I'm happier still.