I own a Strida 5 but few tools, especially not a torque wrench (though I plan to buy one).
How much play should there be in the belt? If I push down on the top of the belt directly above where the lower arm curves, I have perhaps 15cm play. The belt tightening bolt has never been adjusted. It is almost all the way back for maximum tightness. I can loosen the belt tightening bolt, but I cannot move the slider forward or backward without fear of damaging it or the allen key. Surely I need to add a bit of oil.
There is no reason why there should be any difference between folded and unfolded:
the freewheel is fixed on the lower frame tube (actually on the axis linking seat tube and lower tube), so its relative position to the chainring does not vary.
the lower tube defines the rigidity of the transmission and attaching its end to the steering tube does not affect it.
According to me, nothing should prevent you to adust the belt tension on a folded Strida, it is even more convenient than when it’s folded!
I really wonder that you did not recognize the difference, but maybe you are folding your bike another way than me.
The Strida can be folded in two ways, the drawings should explain that:
The drawings are showing the cross section of the three tubes, ~ 25 cms below the ball joint.
Possibility A:
The bottom tube is laying upon the steering tube.
Here is no difference in the belt tension noticeable.
Possibility B:
The bottom tube is clamped between steering and seat tube.
The belt is tensioned more than in unfolded position.
How comes that?
I’m sorry to say, but that’s not fully correct, actually is the freewheel fixed to the seat tube by the rear axle.
(The rear axle is screwed in the seat tube via a left-handed thread from the left side of the seat tube.)
The axle goes through the seat tube, at the right side the axle stump is linked to the bottom tube via the bolt 373 and (important) the cup spring 363.
There is some (not much) backlash between the axle stump and the bottom tube, the high spring force of the cup spring is pressing bottom and seat tube strongly together in unfolded position.
If the Strida is folded to fold type B the cup spring will be compressed and then the bottom tube can be moved sideways - I believe that is the only reason why Ming cycle had to use such a spring.
Would a usual washer be used at that position it would not be possible to fold the Strida to fold type B.
A nearer look makes it visible, actually the position of beltwheel and freewheel together have changed and the distance between the outer edges of the diameters has increased - the result is higher belt tension.
The drawings below are naturally disproportionated.
Please note that for the two pics below neither the cam (tripod) nor the bike frame was moved, only the bottom tube was moved from folding type A to B.
Bigger picture details:
Here the different angles of belt and tubes:
Dear Bietrume, I’m used from work that the engineers do not trust their workers, but I hoped that you know me better…
Honestly I answered from my office and did not check on the bike, so it was a typical engineer answer , purely theoretical and not validated by the practice
So, I totally agree with your explanation. Anyway, it would have been more logical to have the rear axle fixed on the lower tube, but there is probably a good reason for this.
But I guess that this misalignment could eventually take place in unfolded position when putting some effort on the pedal, causing some problems such as belt twisting??
By the way the recommendation of adjusting the belt tension when unfolded is missing in the manual .
I just knew the answer for this behaviour, but I did not know if anybody noticed
Very good to have your agreement now, thank you!
I’m sure that Sara would appreciate improvement suggestions, maybe we should collect them?
My suggestion is to collect improvements, discuss the list in a new post and then send it to Ming cycle.
, purely theoretical and not validated by the practice 
.