Can prophecies for good fail to materialize

The following is a question on a well known opinion of Rambam from an explicit pasuk

 

Rambam famously holds (hakdamah to peirush hamishnah) that the only way to test a new prophet's mettle is if he foretells good and it fails to materialize (with a carveout for personal communications, eg Yaakov being concerned that sin would disturb Gods past positive prophecy to him); if he foretells bad and it doesnt come to pass that doesnt disprove his claim to prophecy as HaShem relents should people repent (eg Nineveh). 

Rambam says that it has got to be the case that there is a prophecy impervious to future actions as otherwise we'd never be able to determine a prophets credibility. Rambam reads Yirmiyahu 28:7-9 as supporting this conception, Rashi there explains the passage the same way in the name of R Tanchuma. 

Don Yitzchak Abarbanel (ad loc) as well as R Yosef ibn Kaspi (Devarim 18:21) disagree, holding that just as a prophecy for bad can go unfulfilled due to an improvement in behavior, likewise, a prophecy for good can be left unfulfilled due to demerits accrued by the intended recipient(s) (see ibn Kaspi there for how to determine the veracity of a new prophet). They both interpret the aforementioned passage in Yirmiyahu as saying that when you've got two prophets prophesying contradictory prophecies the only way to see which prophet is prophesying truthfully is by the outcome, whichever way it goes (with presumably no material change in circumstances). 

In Ki Savo, in the Tochacha, theres a verse (28:68) that goes as follows: וֶהֱשִׁיבְךָ י"י מִצְרַיִם בׇּאֳנִיּוֹת בַּדֶּרֶךְ אֲשֶׁר אָמַרְתִּי לְךָ לֹא תֹסִיף עוֹד לִרְאֹתָהּ וְהִתְמַכַּרְתֶּם שָׁם לְאֹיְבֶיךָ לַעֲבָדִים וְלִשְׁפָחוֹת וְאֵין קֹנֶה, seemingly disproving Rashi/Rambam and cohering with Abarbanel/ibn Kaspi, as it has Moshe saying that HaShem will go back on his previously communicated promise for good (not to bring them back to Egypt) should Israel sin. 

Surprisingly, Ibn Kaspi doesnt bring proof to his previously stated position from this verse. Abarbanel too (ad loc) comments that ships are specified so as not to conflict with the previous prophecy of not being led back to Egypt on the land route taken when they first exited (see Targum yonasan that specifies that to the contrary the ships will sail on the yam suf in the specific place that they traversed where they were promised that they'd never need to see again!), which, aside for its inherent difficulty as an acceptable explanation (Gods point was that they wouldn't ever need to return to mitzrayim proper, not that they would return but not by retracing their exact steps), is entirely unnecessary according to his opinion, as a prophecy for good can too go unfulfilled.

In any event it seems to present a serious difficulty for R Tanchuma, Rashi and Rambam. 

One of the limitations placed on a king is not to possess unnecessary horses. The pasuk reads (Devarim 17:16) רַק לֹא יַרְבֶּה לּוֹ סוּסִים וְלֹא יָשִׁיב אֶת הָעָם מִצְרַיְמָה לְמַעַן הַרְבּוֹת סוּס וַי"י אָמַר לָכֶם לֹא תֹסִפוּן לָשׁוּב בַּדֶּרֶךְ הַזֶּה עוֹד, ostensibly saying that pursuing the gratuitous acquisition of horses will probably lead the nation back to Egypt (as a world leader in all things equine), which is problematic as God has already disallowed our return to mitzrayim. This pasuk is apparently referring back to when Israel was by the yam suf, which has Moshe proclaiming (Shemos 14:13) וַיֹּאמֶר מֹשֶׁה אֶל הָעָם אַל תִּירָאוּ הִתְיַצְּבוּ וּרְאוּ אֶת יְשׁוּעַת י"י אֲשֶׁר יַעֲשֶׂה לָכֶם הַיּוֹם כִּי אֲשֶׁר רְאִיתֶם אֶת מִצְרַיִם הַיּוֹם לֹא תֹסִפוּ לִרְאֹתָם עוֹד עַד עוֹלָם, which is the only previously recorded mention of it. The issue is that that pasuk doesn't seem to be an issur but rather a havtacha that they needn't worry as theyll emerge victorious and never need to return. R Avraham ibn Ezra addresses this (on the shofetim verse): לא תוספון – מצוה היתה, ולא נכתבה.

ויש אומרים: מטעם לא תוסיפו לראותם עוד (שמות י"ד:י"ג). וזאת דרך אחרת.
positing an unrecorded mitzvah of not returning as the beshalach verse cannot be the referent as it is not an issur. 

Rashi to Shemos, per Mizrachi, parses the verse in a way that indeed yields a prohibition, in accordance with Chazals statement (eg Mechilta dRebbi Yishmoel ad loc, PT Sukkah 5:1) that the Torah records it three times. 

Ditto for Lekach Tov (11th century) to the pasuk in the tochacha.

The only one of the three mentions that is easily and naturally read as an admonition—in the world of peshat—is the pasuk in shofetim. 

(Even there see Targum Yonasan's unique reading where he explains it as a precaution against behaviors that'll result in another galus mitzrayim causing God to abrogate his original promise of never returning them to mitzrayim, doing away with any mention of a prohibitory statement.) 

Rambam (sefer hamitzvos lavin 46) quotes the shofetim verse as the source for the prohibition of not returning to mitzrayim. He then cites Chazals statement to the effect of there being three mentions of it, commenting: אע"פ שהנראה מהדברים שהוא ספור באה הקבלה שהוא אזהרה. 

In any event, in Chazals reading it wouldn't be an issue, as they are somehow all conceived of as proscriptive in nature, however, in the world of peshat, with the pesukim being understood predictively, the ki savo verse, at the very least, seems to present a formidable challenge to Rambams well known opinion regarding how to determine a new would-be prophet's bona fides.