Open menu to save: 1s
First save: 23s 
Total S+Q w/ time change: 1:13
[Date is included. Skipping would save 3.3s]
[Included is a ~7 second wait for delay]
[Time change took about 4s. This is pretty reasonable, it's like ~3 to one and ~20 to the other. Could maybe save 1-2s here with a real nice lineup, but not much]

Subsequent saves: 8.5s, so this is the timing to use.

So a full S+Q would be 58.5s on same date change and wait
Assuming a good seed without the wait, we can say ~52s
Assuming no date change and only time, we can say ~49s

Vs manip; 
Average time for an encounter (including time searching) is roughly 16.6s

Assume ~5 extra seconds for final manip 
[Since manipped tenta is probably quicker. A rough estimate of a 5 second difference would be something like, Bike 30 + Walk 5 (6.8s) vs Bike 10 (1.8s)]

So a 1 encounter manip would be ~21.6 	| Saves ~27.4s
2 encounter manip would be ~38.2		| Saves ~11.8s
3 encounter manip would be ~54.8		| Loses ~5.8s


---------- Conclusion

In an ideal world, it's definitely faster to do an extended manip. The problem is that finding an extended manip that lines up that fast isn't exactly easy. It probably requires a full side by side comparison to see what's really the best option, e.g. playing out both seeds (which is an annoyance to check...).

Personally I think it comes down to the difference in the uncalcs for this scenario. E.g. can you always hit the L40 Tento manip? No one is going to be 100% at the execution of hitting the one framer there, and a missed manip costs about 38 seconds (assuming great setup)

However, doing a no s+q has harder catch manip. Even before that, depending on the seed, you may have encounters faster than you that you can't flee from. To me it depends on if there's an extended option that gets rid of these problems from the theory in practice.