If even fewer people donate than expected, that doesn’t undermine the argument. If anything, that suggests the level of support for an education party is even lower.

The author noted that an education party is “unlikely” to be viable. This recognizes that it’s possible some education parties might be able to gain viability even if they don’t meet the requirements observed based on the historical evidence.

The overall amount of money donated is irrelevant to the author’s reasoning, because the historical evidence relied on is about % of voters who are prepared to join or donate. The historical requirement for viability isn’t about the amount of money that a party needs.

The author considers this. The author acknowledges that over 50% support the idea of an education party, but states that historically there are specific kinds of support required (join/donate). That’s why the author thinks the “over 50% support” isn’t enough for viability.

If some of the voters who would donate are not those who would join, that shows it’s possible the combination of voters who would join plus the voters who would donate could exceed 30%. This is why the premises don’t show that an education party is likely to fail.