Let's have a go at splitting some hairs...
Browsing through the Breitkopf score for the Scena (No.2 of Scènes historiques I, Op.25) I am struck by a rather strange tempo marking.
The piece starts out at Tempo di menuetto (fine), then changes to Allegro moderato for the first set of fanfares (interspersed with various stretto/a tempo markings - also fine) but what comes next confuses me. Once the mysterious sul ponticello passage starts (letter C) the tempo is Allegro, poco a poco più, which ought to mean a gradual accelerando. That works throughout this passage, but Sibelius doesn't alter the marking for the concluding march section (Letter H). If we are to follow the letter of the score, therefore, the march should continue to speed up all the way to the final bars. Just before the march he changes from 4/4 to 2/2 which would suggest that the pulse at the beginning of the march ought to be roughly the same as the pulse at the beginning of the sul ponticello passage, but in doubled note values. I don't ever recall hearing a performance that speeds up during the march section (and musically I would find it hard to justify).
The 'original version' (i.e. the fourth tableau of the Press Celebrations Music) also starts at Tempo di menuetto, goes to Allegro moderato for the fanfares, which is maintained (a tempo) for the start of the equivalent of the sul ponticello passage, later speeding up (poco a poco più allegro), reaching a steady new tempo, Allegro molto, 8 bars before the march begins.
Might Sibelius simply have forgotten to mark the point at which the tempo stabilizes? Or is it an error in the published score? (It isn't yet a JSW critical edition, but Breitkopf's scores are pretty reliable on the whole.)