The REDUCE function is powerful, but the blank-accumulator problem is a genuine friction point that spreadsheet users shouldn't have to work around with clunky workarounds. This user has identified a real gap in how REDUCE handles initial values when you want to build arrays cleanly, and their frustration is understandable. The method they landed on, using INDEX to seed the accumulator and then checking equality, works, but it's the kind of fix that makes a formula harder to read, harder to debug, and harder to share with colleagues who might inherit the workbook.
What this reveals is a deeper design tension in modern spreadsheet tools. REDUCE was built for flexibility, but flexibility without clear guardrails can create edge cases that punish users for wanting to do something straightforward. The user's goal is simple: transform each value in a column and return the results in a new column. That should not require a nested IF with an AND check just to avoid a phantom blank or a skipped first value. When a function forces you to choose between an empty first cell and losing your first transformation, the function is asking too much of the user.
The practical lesson here is about knowing when to use REDUCE and when to reach for a simpler alternative. For this specific task, applying the same transformation to every value in an array, MAP or BYROW would give you clean results without the accumulator headache. REDUCE shines when you need to carry forward a running total or combine values in a non-trivial way. Using it to build a transformed array is possible, but the user's experience shows that "possible" and "practical" are not always the same thing. The workaround they built is a testament to their problem-solving skills, but it's also a sign that the tool could be more intuitive.
If you find yourself writing a formula that requires a conditional check on the accumulator just to avoid a blank cell, that is a strong signal to reconsider your approach. Explore MAP, SCAN, or even a simple array formula before committing to REDUCE. The best formula is not the one that technically works, it is the one that a teammate can read six months later and understand in seconds. The user in this thread did the hard work of identifying the problem and finding a solution. The next step is choosing the right tool for the job, not the most clever one.