From 50 reported cases between 2016 and 2026, the CRA won outright in 32 representing a 64 percent win rate. Taxpayers won outright in only 11, or 22 percent. The remaining 7 or 15 percent, were mixed. A claimant who challenges the CRA in court has roughly a one-in-five chance of an unqualified win and a four-in-five chance of losing in whole or in part.

‍

The reasons are remarkably consistent. SR&ED Eligibility drives 32 of the 50 disputes, and within that group, 20 cases turned on a failure of the "Northwest Hydraulic" test. Judgments cite the same recurring shortfalls from this test: no demonstrable technological uncertainty, no systematic investigation, no hypothesis formulation, reliance on trial-and-error rather than scientific method, work that amounts to standard engineering, and inadequate contemporaneous documentation. Other losses involve missing prescribed information on Form T661, failed capital expenditure thresholds, and procedural defects.

‍

We feed this analysis directly into how we structure narratives, capture uncertainty, and assemble evidence  so a claim never has to be litigated.

‍

Mobile Capital Network was the claim preparer for Allegro Wireless Canada Inc., whose appeal became Allegro Wireless v. The Queen, 2021 TCC 27. The decision stands out for two reasons.

‍

First, it is a clean taxpayer victory in a category where the CRA wins overwhelmingly. Of the 32 reported eligibility disputes, taxpayers won outright in only a handful. Allegro is one of them.

‍

Second, it is a software case, a category where claimants historically struggle because reviewers often dismiss iterative coding as routine development. What made the difference at Allegro was not the existence of evidence. Source control history, bug-tracking entries, and iteration logs were already there. What made the difference was how MCN technically positioned that evidence.

‍

Our team's deep technical expertise meant we understood what the developers were actually doing, and why it was hard. We identified that the true source of technological uncertainty was not the application code itself but the interaction with opaque, "black-box" third-party software whose internal behaviour could not be observed or predicted. We articulated the precise nuances of the development effort — the hypotheses being tested, the systematic methodology behind each iteration, and the resulting technological advancement, in the exact terms the "Northwest Hydraulic" test demands. The court saw work that satisfied all five criteria because we framed it that way.

‍

The evidence was there. The technical positioning is what made it win.

‍

Allegro Wireless is the exception that proves the rule. The base-rate truth, established across a decade of decisions, is that litigating an SR&ED dispute is a bad expected outcome — and the strategic consequences extend well beyond legal costs.

‍

The first is timing. SR&ED is a cash-flow program, and once a claim is in dispute, downstream claims are often held up while the matter moves through Notice of Objection, Tax Court scheduling, hearing, judgment, and any appeal. Two to four years of delay is common. For an R&D-intensive business, that delay can be more damaging than the disallowance itself.

‍

The second is risk profiling. CRA reviewers operate against risk-assessment frameworks, and a history of litigation shifts a claimant into a higher-scrutiny segment, meaning more frequent reviews, deeper document requests, and lower tolerance for ambiguity on every claim that follows.

‍

The third is the underlying odds. With 78 percent of cases ending in full or partial loss, the rational strategy is not to get better at fighting the CRA.  It is to submit a claim robust enough that the dispute never starts. Build it once, build it right, and the question of suing the CRA never has to be asked.